Policymaking in Latin America How Politics Shapes Policies by IDB

GOVERNMENT AND PUBLIC INSTITUTIONS

IDB

Policymaking in Latin America is based on the premise that good policies must begin with good institutions, that is, institutions that produce stable social consensus and reliably commit governments over time to a consistent course of action. Looking systematically at the policymaking framework across Latin America it extracts useful lessons about the process that produces good policies. This book is critical reading for anyone interested in reform in this region. —Francis Fukuyama, Bernard L. Schwartz Professor of International Political Economy and Director, International Development Program, The Paul H. Nitze School of Advanced International Studies, Johns Hopkins University Policy analysis has been motivated until now by the question of what governments should do. But governments seldom do what they “should,” leaving the analyst as an ineffectual preacher of virtue. Here is a book that breaks out of this paradigm and proposes a framework to analyze how it is that policy decisions are made. This should lead to an agenda that answers the question of how to decide, not just what to decide. —Ricardo Hausmann, Professor of the Practice of Economic Development, John F. Kennedy School of Government and Director, Center for International Development, Harvard University This volume provides fascinating interpretations of the roles of political competition, legislative structure, judicial independence, and other institutional factors in the making of Latin American economic policy. Policymaking in Latin America makes important contributions to the analysis of the political economy of developing countries, of relevance well beyond Latin America. —Jeffry A. Frieden, Stanfield Professor of International Peace, Department of Government, Harvard University

1300 New York Ave., NW Washington D.C. 20577 www.iadb.org Distributed by Harvard University Press

978-1-59782-061-5

Policymaking in Latin America How Politics

Policies

Ernesto Stein and Mariano Tommasi, editors (with Pablo T. Spiller and Carlos Scartascini)

Policies

Inter-American Development Bank

Policymaking in Latin America: How Politics

What determines the capacity of countries to design, approve, and implement effective public policies? To address this question, this book builds on the results of case studies of political institutions, policymaking processes, and policy outcomes in eight Latin American countries, structured around a common methodology with foundations in game theory and institutional analysis. The result is a volume that benefits from both micro detail on the intricacies of policymaking in individual countries and a broad cross-country interdisciplinary analysis of policymaking processes in the region.

Inter-American Development Bank

DRCLAS HARVARD

David Rockefeller Center for Latin American Studies Harvard University

Policymaking in Latin America How Politics Shapes Policies

Ernesto Stein and Mariano Tommasi Editors With Pablo T. Spiller and Carlos Scartascini

Inter-American Development Bank David Rockefeller Center for Latin American Studies Harvard University

©2008 Inter-American Development Bank 1300 New York Avenue, N.W. Washington, D.C. 20577 Co-published by David Rockefeller Center for Latin American Studies Harvard University 1730 Cambridge Street Cambridge, MA 02138 Copyright © by the Inter-American Development Bank. All rights reserved. No part of this book may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, or by information storage or retrieval system, without permission from the IDB. To order this book, contact: IDB Bookstore Tel: (202) 623-1753 Fax: (202) 623-1709 E-mail: idb-books@iadb.org www.iadb.org/pub The views and opinions expressed in this publication are those of the authors and do not necessarily reflect the official position of the Inter-American Development Bank. Cataloging‑in‑Publication data provided by the Inter‑American Development Bank Felipe Herrera Library Policymaking in Latin America : how politics shapes policies / Ernesto Stein and Mariano Tommasi, editors; with Pablo Spiller and Carlos Scartascini. p. cm. Includes bibliographical references. ISBN: 978-1-59782-061-5 1. Public administration—Latin America—Decision making. 2. Policy sciences. 3. Public policy. 4. Political planning—Latin America. 5. Pressure groups. 6. Political parties. I. Stein, Ernesto. II. Tommasi, Mariano, 1964-. III. Spiller, Pablo. IV. Scartascini, Carlos. V. Inter-American Development Bank. VI. David Rockefeller Center for Latin American Studies. JL959.5 .D45 P84 2007

LCCN: 2007940418

Cover Image: Valérie Crausaz, France, Untitled, 1998, Color etching on Arches paper, 40-15/16 in. × 13-3/8 in., Inter-American Development Bank Art Collection Cover Design: Ultradesigns.com

Contents

Acknowledgments..................................................................... vii FOREWORD........................................................................................ xi Chapter 1. Political Institutions, Policymaking, and Policy: An Introduction . ..........................................1 Pablo T. Spiller, Ernesto Stein, and Mariano Tommasi Chapter 2. Who’s Who in the PMP: An Overview of Actors, Incentives, and the Roles They Play..............................29 Carlos Scartascini Chapter 3. Political Institutions, Policymaking Processes, and Policy Outcomes in Argentina................................69 Pablo T. Spiller and Mariano Tommasi Chapter 4. On the Road to Good Governance: Recovering from Economic and Political Shocks in Brazil.....................111 Lee J. Alston, Marcus André Melo, Bernardo Mueller, and Carlos Pereira Chapter 5. Political Institutions, Policymaking Processes, and Policy Outcomes in Chile.....................................155 Cristóbal Aninat, John Londregan, Patricio Navia, and Joaquín Vial Chapter 6. Political Institutions and Policy Outcomes in Colombia: The Effects of the 1991 Constitution..........................199 Mauricio Cárdenas, Roberto Junguito, and Mónica Pachón

Contents

Chapter 7. Veto Players, Fickle Institutions, and Low-Quality Policies: The Policymaking Process in Ecuador.........243 Andrés Mejía Acosta, María Caridad Araujo, Aníbal Pérez-Liñán, and Sebastián Saiegh Chapter 8. Policymaking in Mexico Under One-Party Hegemony and Divided Government.............................................287 Fabrice Lehoucq, Gabriel Negretto, Francisco Aparicio, Benito Nacif, and Allyson Benton Chapter 9. Political Institutions, Policymaking Processes, and Policy Outcomes in Paraguay................................329 José Molinas, Anibal Pérez-Liñán, Sebastián Saiegh, and Marcela Montero Chapter 10. Political Institutions and Policymaking in Venezuela: The Rise and Collapse of Political Cooperation........371 Francisco Monaldi, Rosa Amelia González, Richard Obuchi, and Michael Penfold References .................................................................................. 419 INDEX ............................................................................................. 465

Contents

Figures Figure 1.1 Political Institutions, Policymaking Process, and Policy Outcomes............................................................... 8 Figure 3.1 Volatility of Economic Policies, Selected Countries, 1970–2003........................................................................................72 Figure 3.2 “Weberianness” of the Bureaucracy in Relation to Human Development in 35 Developing Countries.....................104 Figure 5.1 Trade Liberalization, Chile, 1974–2004...............................161 Figure 5.2 Distribution of Social Transfers by Income Quintile, 1997–2001......................................................................................168 Figure 6.1 Fiscal Balance of Central Government, Colombia, 1905–2005......................................................................................200 Figure 6.2 Inflation in Colombia, 1950–2006.......................................201 Figure 6.3 Number of Lists Running Per Election, Colombia, 1958–2006......................................................................................209 Figure 6.4 Effective Number of Political Parties in the House of Representatives, 1974–2006...................................213 Figure 6.5 Average Number of Legislative Sponsors, Colombia, 1982–2003......................................................................................217 Figure 7.1 Evolution of Presidential Approval Rates and the President’s Coalition in Congress over the Term, Ecuador, 1984–2002......................................................................269 Figure 9.1 Effective Number of Parties in the Chamber of Deputies, Paraguay, 1954–2004................................................347 Figure 9.2 Party Unity, Paraguay, 1994–2004........................................351 Figure 9.3 Success Rate of Legislative Bills, by Administration and Initiator, Paraguay, 1992–2003......................................................356 Figure 9.4 Success Rate of Bills, by Initiator, Administration, and Type of Policy, Paraguay, 1992–2003............................................357 Figure 10.1 Effective Number of Parties in Venezuela, 1958–2005......................................................................................378 Figure 10.2 Volatility in the Lower Chamber of the Legislature in Venezuela, 1963–2005...................................................................379 Figure 10.3 Governance Indicators, Selected Countries, 2004............408

Contents

Tables Table 3.1 Policymaking Capabilities of Latin American Congresses.....85 Table 3.2 Reelection Rates to the Argentine Chamber of Deputies, 1985–2003........................................................................................88 Table 3.3 Congressional Service by Argentine and Costa Rican Legislators...................................................................88 Table 3.4 Reelection Rates in Selected Countries, Various Years...........89 Table 3.5 Supreme Court Justices’ Average Tenure, Selected Countries, 1960–1990 ...................................................101 Table 6.1 Constitutional Presidential Powers, Colombia, 1958–2004......................................................................................206 Table 6.2 Legislative Database, Colombia, 1982–2003.........................222 Table 6.3 “Determinants of Legislative Success Rates,” Colombia, 1982–2003 (Logit Model).............................................................224 Table 6.4 Characterization of Policies Before and After the 1991 Constitution: Outer Features, Colombia............................239 Table 7.1 Ecuadorian Presidents and Congressional Support, 1979–2005......................................................................................245 Table 7.2 Bills Initiated and Approved, by Initiator and Type, Ecuador, August 1979–April 2004................................................259 Table 7.3 Initiation and Approval of Bills (Logistic Regression), Ecuador . .......................................................................................271 Table 7.4 Initiation and Approval of Bills (Logistic Regression), Ecuador..........................................................................................273 Table 8.1 Evolution of the Policymaking Process in Mexico, 1950–2006......................................................................................291 Table 8.2 The Partisan Balance of Power between the President and Congress, Mexico, 1952–2003...............................................302 Table 8.3 Corporatist Affiliation of PRI Deputies, Mexico, 1964–2003......................................................................................306 Table 8.4 Legislative Bills Introduced in and Passed by the Chamber of Deputies, Mexico, 1991–2006...........................310 Table 9.1 Passage Rate and Presidential Veto of Legislative Bills, by Initiator and Policy Area, Paraguay.........................................358 Table 9.2 Comparison of the Policymaking Process in Different Historical Periods, Paraguay, 1954–2003.....................................368

Acknowledgments

he initial inspiration for this project came from the work on Argentina by two of the editors of this book, Pablo Spiller and Mariano Tommasi, reflected in their book The Institutional Foundations of Public Policy in Argentina: A Transactions Cost Approach (Cambridge University Press, 2007). In that work, Spiller and Tommasi developed a methodology that, with some refinements and adaptations, became the basis for the conceptual framework used in this project, in this case applied to a much larger set of countries in Latin America. The book is part of a larger agenda on political institutions and policy outcomes in Latin America being carried out by the Research Department of the Inter-American Development Bank (IDB), which includes among other products the 2006 Economic and Social Progress Report, The Politics of Policies. The country chapters in this book were produced by eight interdisciplinary country teams under the auspices of the IDB Latin American Research Network, and the coordination of the four editors. The process of writing the book was highly interactive, with ample opportunities for cross-fertilization among the authors of the country chapters, as well as frequent give and take (in both directions) between these authors and the project coordinators. In addition to the authors of the country chapters in the book, we are also indebted to Mario Bergara, and to Eduardo Moron and Cynthia Sanborn, who participated actively in this interactive process, but whose studies on Uruguay and Peru were ultimately not included in this volume for reasons of space. Two seminars organized by the IDB, the initial one in Mexico City in October 2003 and the second one at Harvard University in May 2004, were very important focal points in this interactive process. For their

viii

Acknowledgments

help in organizing these seminars, as well as for their invaluable support during the whole process (from the initial call for proposals to the completion of the final drafts), we want to thank Norelis Betancourt and Raquel Gomez. The second of these seminars was co-organized with Jeff Frieden, Stanfield Professor of International Peace at Harvard University, who helped us put together the seminar back to back with the prestigious biannual conference on Political Institutions and Economic Policy (co-organized with Ken Shepsle). Jeff assembled an all-star roster of political scientists (including, in addition to himself, Jim Alt, Robert Bates, Adam Przeworski, and Ken Shepsle) and asked them to provide feedback on different aspects of the overall project. Their invaluable contributions were a key input for the project. Beyond his role in the organization of this seminar, Jeff has provided constant encouragement and feedback, and we owe him a huge debt of gratitude. Similarly, we are very grateful to Fernando Carrillo from the IDB Office in Europe, and to Javier Santiso and BBVA, who hosted a workshop in Madrid so that we could disseminate our findings in Europe, and receive feedback from European scholars. We would also like to recognize our colleagues at the Research Department of the IDB and at Universidad de San Andrés, for their support, their encouragement, and their feedback. Within the IDB, we want to especially recognize the support received from Guillermo Calvo, at the time the IDB’s Chief Economist, and his Principal Advisor, Eduardo Lora. Without their support, this project would not have been possible. Tommasi acknowledges the support of the John Simon Guggenheim Memorial Foundation. Apart from the authors of the chapters and those already recognized above, many others deserve recognition for their valuable comments at different stages of the process. This includes Carlos Acuña, Manuel Alcántara, Martín Ardanaz, Xosé Carlos Arias, Guillermo Cardoza, John Carey, Fernando Carrillo-Flores, Gerardo Della Paolera, Guillermo de la Dehesa, Carlos Elizondo, Jose Luis Escrivá, Ricardo Hausmann, Jacint Jordana, Marcelo Leiras, María Victoria Murillo, Alejandro Nuet, Mauricio Olivera, Charles Oman, Ludolfo Paramio, Santiago Pombo, Joan Prats, Jessica Seddon Wallak, Thierry Verdier, Andrés Velasco, Lawrence

Acknowledgments

Whitehead, and two anonymous referees. Participants in seminars at LACEA, LACEA Political Economy Group, IDB, and the Universidad de San Andrés, in which some of the papers were presented, should be recognized as well. Ideas become successful books thanks to capable editorial and administrative support. For their invaluable support in this area, we would like to acknowledge Rita Funaro, María Helena Melasecca, Mariela Semidey, John Smith, and Luis Vélez. For pulling everything together we are most grateful to the main editor, Nancy Morrison, as well as the publications team at the IDB, led by Rafael Cruz and including Michael Harrup and Joanne Blake. Finally, this book would not have been possible without the exceptional contributions of Heather Berkman, who provided research assistance, managed the interaction with the country teams late in the process, and carefully followed the production process through the later stages of the project.

Foreword

atin America is engaged in a constant search for policies to accelerate growth, reduce poverty, and otherwise further its economic and social progress. In its journey on the path to development, it has been offered, and often followed, celebrated road maps designed to speed it along in this tortuous trek. The state-run inward-looking policy package of the postwar era and the liberalization of the Washington Consensus were each hailed as tickets to sustainable, equitable development. Instead, they produced mixed results and ultimately fell short of the region’s goals and expectations. Were these policies flawed? Was something missing? To date, most of the discussion has been an exchange between those who argue that the transformation of the state has been incomplete and more reform is needed, and those who oppose reform and attribute the poor results to them. This book suggests an alternative view: the problem lies less in the policies than in the process behind these policies. Public policies are not simply items on a menu that policymakers pick and choose. Rather, they are cooked up by numerous political actors and must then be implemented and sustained over time. Each of these political actors at each stage of the process brings his personal interests to the table and is pressured by others in the process with their own respective interests. The quality of the outcomes in the policymaking process depends as much on how these different actors interact as on the merits of the policy being promoted. Thus, differently put, the issue is not only with the nature of the interventions pursued by the state (“Producing”, “Regulating”, or “Distributing”), but also with the institutions that constitute the state: their incentives, the rules that govern their day-to-day functioning, and their accountability.

xii

Foreword

This book focuses on the political dimension of the public policymaking process. It analyzes the results of a comparative study of political institutions, policymaking processes and policy outcomes in eight Latin American countries, which together account for nearly 80 percent of the population and 90 percent of the GDP of the region. Each country chapter provides a detailed description of the electoral, legislative, executive, judicial, and bureaucratic institutions of the country and analyzes how these political institutions interact as they go about the business of making economic policy. By comparing both the processes and the policy outcomes in these countries, the book goes a long way toward explaining why things work the way they do in different contexts. Policymaking in Latin America: How Politics Shapes Policies comes at an opportune moment in Latin America. Political institutions and the political rules of the game are under scrutiny in many countries and in some of these countries are currently subject to change. Constitutional reforms are on the agenda in some countries, the role of the state is being reviewed and the balance of political power may be shifting. This book offers a wealth of information that can serve as valuable input for these discussions. The book is of great relevance to policymakers to the extent that it can help avoid costly mistakes of the past, on one hand. On the other, because it points out an issue at times underestimated by economists: that the interactions between actions and institutions in the policymaking process are as important as the nature of the policies being promoted in determining outcomes. Santiago Levy Chief Economist and General Manager Department of Research Inter-American Development Bank

1 Political Institutions, Policymaking, and Policy: An Introduction Pablo T. Spiller, Ernesto Stein, and Mariano Tommasi

“The political process that underpins economic reform is no less important than the reform’s content, and perhaps even more so.” —Dani Rodrik

his book is about the political economy of policymaking in Latin America. It is written for an audience of development practitioners, economists, and political scientists, as well as for students training in those disciplines with an interest in Latin America. The objects of study are institutions, political behavior, and public policies. In addressing the links between political institutions, policymaking processes, and policy outcomes in a variety of countries, the book builds bridges between the concerns of economists about the effects of public policies on economic and social outcomes, and the concerns of political scientists about the workings of democratic institutions and political outcomes. Public policies have always been a central concern for development economists and other development practitioners. In fact, the work of the development community has been dominated by the search for policies that would help accelerate growth, reduce poverty, or achieve other noble economic and social goals. As a result of this search, Latin America has been repeatedly influenced by various waves of policy advice, from the

Pablo T. Spiller, Ernesto Stein,

and

Mariano Tommasi

state-run inward-looking development of the postwar era to the liberalization of the Washington Consensus in the 1990s. In both these cases, the results of these efforts have been diverse, and somewhat disappointing. This book suggests a change of focus in this search. It starts from the premise that public policies are not simply objects of choice for a social planner trying to maximize the welfare of the population. Rather, public policies emerge from a decision-making process that involves a multiplicity of political actors who interact in a variety of arenas. Thus the studies in this book look beyond the specific content of policies and focus on the policymaking process (PMP): that is, on the critical process that shapes policies, carries them forward from idea to implementation, and sustains them over time. This focus is grounded in the belief that the potential of policy prescriptions to deliver better outcomes ultimately depends, to an important extent, on the quality of the policymaking process through which policies are discussed, approved, and implemented. Policymaking processes affect the type and content of policies that are adopted. But these processes may also imprint some common characteristics on public policies. These characteristics may be as important as the policy content itself. Policymaking processes can contribute to policy stability or lead to large policy swings; they may facilitate policy adaptability or lead to excessive rigidity; they can produce policies that promote either the public welfare or private interests; they can affect the quality of policy implementation and enforcement. In short, policymaking processes can affect the nature and the quality of public policies in a variety of dimensions. In the Latin American democracies studied in this book, the process of adopting and implementing public policies occurs in political systems in which a variety of actors participate, ranging from the president to voters in small rural communities, and including members of congress, judges, public opinion leaders, and business groups. The complex interaction among these actorsâ&#x20AC;&#x201D;the subject of this bookâ&#x20AC;&#x201D;is influenced by the institutions and political practices of each country. These institutions affect the roles and incentives of each of the actors, the characteristics of the arenas in which they interact, and the nature of the transactions they engage in.

Political Institutions, Policymaking,

and

Policy: An Introduction

The emphasis on the role of domestic political institutions as the explanatory variables of interest in this volume represents somewhat of a departure from the focus of much of the earlier work on Latin American political economy, which placed relatively more emphasis on the impact of foreign economic pressures and powerful special interest groups. That earlier emphasis was perhaps justified by the conditions prevailing in the 1960s and 1970s, when most of the region was characterized by undemocratic regimes dominated by powerful interest groups in a highly protectionist setting, where favored actors were often able to extract sizable rents and where foreign actors were practically the only meaningful counterweight to the ruling elites.1 These conditions are long gone (at least in most of the region), and the countries of Latin America certainly deserve an openminded analysis of the laws of motion of their political economies, and a more serious investigation of their formal political institutions.2 This book builds upon the efforts of important recent work on political institutions in Latin America. There are some excellent books on comparative politics of Latin America, focusing on specific institutional features in various countries. These include edited volumes such as Mainwaring and Scully (1995a) on party systems, Mainwaring and Shugart (1997a) on constitutional and partisan powers of the president, Carey and Shugart (1998b) on executive decree authority, and Morgenstern and Nacif (2002) on legislative politics. This book benefits from the insights of those volumes and of related work, but takes a more general look at the process of policymaking, including a wider set of actors and their interactions, and links these interactions with the resulting policy features. The book presents the results of a comparative study of political institutions, policymaking processes, and policy outcomes in eight Latin 1

In the words of Barbara Geddes in her survey on the study of politics in developing countries, â&#x20AC;&#x153;When authoritarian governments ruled most developing countries, few political scientists interested in these countries paid much attention to the development of theories of democratic politicsâ&#x20AC;? (Geddes 2002, p. 343). 2 After several decades in which most Latin American countries frequently switched between democratic and military government, the countries of the region gradually returned to democratic rule in the 1980s.

Pablo T. Spiller, Ernesto Stein,

and

Mariano Tommasi

American countries, which comprise nearly 80 percent of the population of Latin America and nearly 90 percent of its GDP. The countries are Argentina, Brazil, Chile, Colombia, Ecuador, Mexico, Paraguay, and Venezuela.3 While similar in some dimensions (they all have presidential systems, for example, and most of them share important aspects of their cultural heritage), they offer substantial variation in terms of country size and level of development, as well as an array of political and institutional features. As the reader can verify by looking at the country cases, they also vary substantially in terms of the quality of their public policies, and thus offer a good basis for studying the impact of political institutions on policy outcomes. The country chapters in this book are not just a collection of related papers. They are the result of a truly collective comparative enterprise in which various multidisciplinary country teams of local scholars, former policymakers, and foreign experts—interacting frequently and intensively—conducted a series of country studies structured around a common framework. As coordinators of a complex comparative project such as the one presented in this book, we had to confront the question of how much guidance to provide the authors of the country chapters in terms of the methodological approach to study the links among political institutions, policymaking processes, and policy outcomes in their own countries. A minimalist approach would have entailed giving them freedom to tackle each study as they saw fit, stressing the factors that were particularly relevant to their own countries. At the other extreme, we could have imposed a rigid common methodology to maximize comparability across the country cases. We opted for an intermediate approach. We provided a guiding framework, inspired by the work of Spiller and Tommasi (2003) on Argentina. The framework, which emphasizes the role of cooperation among the political actors that participate in the PMP,

The country chapters were written as part of an IDB Research Network project on “Political Institutions, Policymaking Processes and Policy Outcomes.” In addition to these studies, the original project also included studies on Peru (Moron and Sanborn 2006) and Uruguay (Bergara et al. 2006).

Political Institutions, Policymaking,

and

Policy: An Introduction

provided a common lens with which to look at policymaking processes, their institutional determinants, and their impact on policies.4 While we asked the authors to take this framework as a point of departure, we also encouraged them to combine it when necessary with other approaches, and to add as many dimensions as needed in order to capture the essence of the policymaking process in their own countries, and its impact on public policies. The country chapters pay extremely detailed and systematic attention to political institutions. Each of them provides a comprehensive description of the electoral, legislative, executive, judicial, and bureaucratic institutions of the country, which in itself is of substantial value. Each chapter also goes on to present an integrative interpretation of how these institutions interact (together with voters and interest groups) in the making of economic policy. The result demonstrates the importance of going beyond simple partial equilibrium assertions about the impact of, say, presidential and parliamentary systems, or independent central banks, and of moving toward a more integrated analysis of how various components of the political-institutional environment interact. Read together, the country studies provide a fascinating interpretation of the role of political competition, electoral institutions, legislative structure, judicial independence, and other crucial factors in the making of Latin American policies. We hope the reader will come away with an appreciation for the complexity of policymaking processes in these countries, and of the central importance of understanding these processes well as a prerequisite for venturing into the challenging territory of giving policy and institutional advice. In this introductory chapter, we summarize the guiding framework and briefly discuss some key insights derived from the country studies.

This framework, which draws from a rich literature on political economy both in economics and in political science, was provided to the country teams at the beginning of the IDB Research Network project on â&#x20AC;&#x153;Political Institutions, Policymaking Processes and Policy Outcomes.â&#x20AC;? See Spiller, Stein, and Tommasi (2003) and references there, as well as in chapter 2 of this book.

Pablo T. Spiller, Ernesto Stein,

and

Mariano Tommasi

A Guiding Framework

The objective of this book is to contribute to the understanding of the determinants of public policies. For example, we want to understand why some countries are able to implement policies that are stable over time, yet are flexible enough to adapt to changing economic conditions, while other countries tend to change policies whenever the political landscape changes, or must resort to highly inflexible and inefficient rules to give policies some stability and credibility. In order to understand these policy outcomes, we focus on the process of discussing, approving, and implementing public policiesâ&#x20AC;&#x201D;the policymaking process (PMP)â&#x20AC;&#x201D;and its grounding in political institutions and practices. The methodological lens we suggested for these studies was based on the idea that several important features of public policies depend crucially on the ability of political actors to achieve cooperative outcomes: that is, their ability to strike political agreements and enforce them over time. In a sense, the whole game of democracy is based on agreement and cooperation at a deeper level: that of respecting the rules of the game and letting other people rule, if that is the votersâ&#x20AC;&#x2122; choice and the outcome of the electoral process.5 In our analysis, within the democratic rules of the game, intertemporal agreement is the mechanism by which political power of the incumbent is not abused: that is, the mutual agreements over time prevent the prevalence of policies that favor the dominant actors of the moment and ignore others.6 In environments that facilitate political agreements, policymaking will be more cooperative, leading to public policies that are more effective, more sustainable, and more flexible in responding to changing economic or social conditions. In contrast, in settings where cooperative behavior is harder to develop

See, for instance, Weingast (1997), Wantchekon (2000), Przeworski (2005), and Fearon (2006). 6 While our discussion focuses mostly on democracies, the framework used in this volume can be also applied to cases of dictatorship. The players may be different (the armed forces may play a big role while the legislature may be weak or nonexisting), but the logic of the framework remains the same. The discussion in chapter 9 of the Stronismo in Paraguay is a case in point. 5

Political Institutions, Policymaking,

and

Policy: An Introduction

and sustain, policies will be either too unstable (subject to political swings) or too inflexible; policies will be poorly coordinated; and there will be little investment in building up long-term capabilities.7 An important question within this framework is whether the workings of the PMP tend to facilitate or hinder cooperative outcomes in policymaking. As we will see in more detail below, the theory of cooperation in repeated games provides some useful insights regarding the type of features of the PMP that may foster cooperation. These include the number of key political actors, their time horizons and the frequency of their interaction, the nature of the arenas in which they interact, and the availability of enforcement mechanisms that bind them to their commitments. The authors of the country chapters were asked to keep these factors in mind when analyzing the PMP of their respective countries. The workings of the PMP, in turn, are determined (to some extent) by the political institutions in place in each country, such as the presidential/parliamentary nature of the government, the electoral rules in place, the federal structure of the country, and the existence of an independent judiciary. In studying the connection of institutional variables with the workings of the PMP, we asked chapter authors to take a systemic “general equilibrium” approach. As we expected from our theoretical priors and from the pilot work we had done on Argentina, the chapters demonstrate that the characteristics of countries’ PMPs do not depend on a single institutional factor, but rather on the interaction among a number of factors—a point we develop in more detail below. The rest of this chapter traces the steps linking political institutions to policymaking processes to policy outcomes. These are the components and steps we requested from the country studies. These components and interconnections are illustrated in figure 1.1. The figure and Rigidity arises when political actors do not trust their opponents, and prefer to tie their opponents’ (and perhaps their own) hands, rather than allow for political discretion. If there are multiple policymaking actors, lack of cooperation is analogous to lack of coordination. It is a standard result of repeated games that in noncooperative environments, agents are less willing to pay short-term cost for future benefits: that is, they are less willing to invest. 7

Pablo T. Spiller, Ernesto Stein,

and

Mariano Tommasi

Figure 1.1. Political Institutions, Policymaking Process, and Policy Outcomes

Characteristics of public policies

Policymaking process

Functioning of political institutions (rules of policymaking game)

Features of specific policy issues

General equilibrium interactions

Basic institutions and history

Source: Spiller and Tommasi (2003).

discussion that follow work backwards, starting from a definition of the dependent variable.8 The Dependent Variable: Public Policies

Policies are complex undertakings. Bringing any particular “policy reform” to fruition is a process that involves multiple actors through many stages of the policy process. It requires specific responses from economic and social agents, and therefore necessitates several forms of cooperation and positive beliefs about the durability and other properties of the policy. That is, policies require a good deal more than a magical moment of special politics to introduce “the right policy” in order to produce effective results. A universal set of “right” policies does not exist. Policies are contingent responses to underlying states of the world. What might work at one point in time in a given country might not work in a differ-

We use “steps” from political institutions to policymaking processes and from policymaking processes to features of policies. This schematic bears some resemblance to the notions of structure, politics, and policy of the comparison of the United States and Japan edited by Cowhey and McCubbins (1995). 8

Political Institutions, Policymaking,

and

Policy: An Introduction

ent place at another time. In some cases, particular characteristics of policies or details of their implementation might matter as much as the broad policy orientation in bringing up the desired outcomes. For instance, Dani Rodrik (1995) analyzed six countries that implemented a set of policies that shared the same generic title of “export subsidization,” but had widely different degrees of success. Rodrik relates success to such features as the consistency with which the policy was implemented, which office was in charge, how this policy was bundled (or not) with other policy objectives, and how predictable the future of the policy was. One important characteristic of policies that has been widely recognized in recent work in economics is policy credibility. The effects of policies on the final economic and social outcomes of interest depend on the actions and reactions of economic and social agents, who take into account their expectations about the future of the policies in question before deciding on their responses. As Rodrik (1989, p. 2) explains, in reference to trade reform, “it is not trade liberalization, but credible trade liberalization that is the source of efficiency benefits. The predictability of the incentives created by a trade regime, or lack thereof, is generally of greater importance than the structure of these incentives. In other words, a distorted, but stable set of incentives does much less damage to economic performance than an uncertain and unstable set of incentives generated by a process of trade reform lacking credibility.”9 For those reasons, in the background work for this book, we suggested that country authors pay special attention not only to the content of public policies, but also to the characteristics of those policies, such as their predictability, their adaptability to changing economic circumstances, and other related qualities. Unlike the content of policies, such features have the advantage of allowing comparisons across multiple policy domains. In addition, they fit very naturally within the proposed theoretical framework. Consider the example of a hypothetical country with two similarly sized parties that are highly polarized, and with elecFor models studying the effects of policies of uncertain duration in several economic contexts, see Calvo (1996, section V) and Calvo and Drazen (1998). 9

Pablo T. Spiller, Ernesto Stein,

and

Mariano Tommasi

toral rules that tend to endow the government with a legislative majority. In such a setting, if the PMP does not facilitate cooperation, as parties alternate in power, the content of some policies may be shifting all the time. Yet features such as policy stability (or, in this case, volatility) will be a constant as long as the institutional setting does not change. We list below some of the policy characteristics (“outer features”) upon which we suggested that authors place their focus. The list, of course, is not taxonomic, and country authors were encouraged to highlight additional features of policies that were considered to be particularly relevant in their own countries. (Several of them focused on the content of policies, as well.) Policy Characteristics (“Outer Features”)

Policy Stability. Some countries seem capable of sustaining policies over time. In other countries, policies are frequently reversed, often at each minor change of political winds (whether a change in administration or even a change in some key cabinet member or senior bureaucrat). By stable policies we do not mean that policies cannot change at all, but rather that changes tend to respond to changing economic conditions or to failure of previous policies, rather than to changes in political administrations, parties, or coalitions. In countries with stable policies, changes tend to be incremental, building upon achievements of previous administrations, and tend to be done through consensus and compromise. In contrast, volatile policy environments are characterized by large swings and by lack of consultation with different groups in society. We associate policy stability with the ability of political actors to strike and enforce intertemporal agreements that allow certain fundamental policies (“Politicas de Estado”) to be preserved beyond the tenure of particular officeholders or coalitions. Thus the notion of policy stability is closely linked to the notion of policy credibility.10

This notion of policy stability is also closely related to the notion of resoluteness in Cox and McCubbins (2001). 10

Political Institutions, Policymaking,

and

Policy: An Introduction

Policy Adaptability. It is desirable for countries to be able to adapt policies to changing economic conditions and to change policies when they are obviously failing. However, governments sometimes abuse the discretion to adapt policies by adopting opportunistic, one-sided policies that are closer to their own preferences or those of narrow constituencies. This can result in policy volatility, as policies may shift back and forth as different groups alternate in power. In political environments that are not cooperative, political actors sometimes attempt to limit such opportunism by resorting to fixed policy rules and mechanisms that are difficult to change.11 This has the benefit of limiting policy volatility, but at the cost of reducing adaptability. In other cases, political systems tend to generate gridlock, making it difficult to achieve change, a point emphasized in the literature on veto players (Tsebelis 1995, 2002) and on divided government. Whatever the reason, countries with low policy adaptability will be unable to respond adequately to shocks, and may get stuck in bad policies for extended periods of time. Coordination and Coherence. Public policies are the outcome of actions taken by multiple actors in the policymaking process. Ideally, different agents acting over the same policy domain should coordinate their actions to produce coherent policies. However, this is not always the case. In some countries on certain issues, policymaking involves a large number of agencies that do not communicate adequately with one another, leading to what Cox and McCubbins (2001) have called â&#x20AC;&#x153;balkanizationâ&#x20AC;? of public policies. Lack of coordination often reflects the noncooperative nature of political interactions. It may occur among different agencies within the central government; between agencies in the central government and others at the regional or municipal level; or even among agents that operate in different stages of the policymaking process (such as when the complications that the bureaucracy might face during the

Some of the countries studied here, such as Brazil and Colombia, have embedded important policy dimensions such as pension policies or intergovernmental transfers into the constitution. 11

Pablo T. Spiller, Ernesto Stein,

and

Mariano Tommasi

implementation phase of a given policy are not taken into account during the design and approval stage of policymaking). Quality of Implementation and Enforcement. A policy could be very well designed, sail through the approval process unchanged, and yet be completely ineffective if it is not well implemented and enforced. In many countries in Latin America, the quality of implementation and enforcement is quite poor. This is associated in part with the lack of capable and independent bureaucracies, as well as the lack of strong judiciaries. To an important degree, the quality of implementation and enforcement will depend on the extent to which policymakers have incentives and resources to invest in policy capabilities. Public Regardedness. This dimension, suggested by Cox and McCubbins (2001), refers to the extent to which policies produced by a given system promote the general welfare and resemble public goods (that is, are public regarding) or whether they tend to funnel private benefits to certain individuals, factions, or regions in the form of projects with concentrated benefits, subsidies, or tax loopholes.12 We have been referring to the features highlighted above as â&#x20AC;&#x153;common featuresâ&#x20AC;? of public policies; although the same features tend to be present in most policy domains, not all policies are expected to have the exact same features. Differences in the nature of the policy issues themselves, in the actors that participate in the respective sector-specific PMPs, and in the arenas where the policymaking game is played may lead to different features across policy areas. As will be discussed below, several of the country chapters take advantage of these differences across sectors, which provide an interesting source of cross-sectional variation within a particular country.

To the extent that those favored by private regarding policies tend to be the members of the elite, who are the ones who have the economic and political clout to skew policy decisions in their favor, private regardedness could relate to inequality. 12

Political Institutions, Policymaking,

and

Policy: An Introduction

The Policymaking Process

The process of discussing, approving, and implementing public policy is referred to as the policymaking process (PMP). In democratic systems such as those in Latin America, these processes play out on a political stage featuring a variety of political actors (or players, in the language of game theory). Players in this game include official state actors and professional politicians (presidents, party leaders, legislators, judges, governors, bureaucrats), as well as business groups, unions, the media, and other members of civil society. These actors interact in different arenas, which may be formal (such as the legislature or the cabinet) or informal (the street), and may be more or less transparent. The PMP can be understood as a process of bargains and exchanges (or transactions) among political actors. Some of these exchanges are consummated on the spot or instantaneously (they are spot transactions). In many other cases, current actions or resources (such as votes) are exchanged for promises of future actions or resources (they are intertem‑ poral transactions). The type of transaction that political actors are able to engage in will depend on the possibilities provided by the institutional environment. Issues of credibility and the capacity to enforce political and policy agreements are crucial for political actors to be able to engage in intertemporal transactions, and to cooperate more generally. The behavior of political actors in these exchanges, and the nature of the exchanges themselves—such as support for the government in a crucial policy issue in exchange for a job in the public bureaucracy, or support for reform in a particular policy area in exchange for concessions in a different policy area—depend on the actors’ preferences, on their incentives, and on the constraints they face. They also depend on the expectations various actors have regarding the behavior of other players. These interactive patterns of behavior constitute what in the parlance of game theory are called equilibria. Thus the characteristics of public policies depend on the equilibrium behavior of these actors in the policymaking game. The behavior of political actors in the policymaking process—as shaped by their roles, incentives, and constraints—will depend, in turn,

Pablo T. Spiller, Ernesto Stein,

and

Mariano Tommasi

on the workings of political institutions (such as congress, the party system, or the judiciary) and on more basic institutional rules (such as electoral rules and constitutional rules) that determine the roles of each of the players, as well as the rules of engagement among them. Policymaking processes are complex; multiple actors with diverse powers, time horizons, and incentives interact in various arenas with diverse rules of engagement. For these reasons, it is not possible to understand these processes fully by focusing on a few institutional characteristics. This level of complexity requires a more systemic approach. Such a systemic view can be accomplished only by means of detailed country studies, which take into account a variety of key institutions and their interaction, as well as historical and cultural legacies.13 This is the reason why we assembled this comparative project, where teams of distinguished international and local scholars, guided by a common methodological lens, and interacting intensely with one another and with the project coordinators, immersed themselves into the details of political institutions, policymaking processes, and policy outcomes in each of the countries under study. One very important building block in each of the studies was a careful characterization of the workings of the PMP, which the authors approached by answering questions such as the following: ❚ ❚ ❚ ❚

Who are the key actors that participate in the PMP? What powers and roles do they have? What are their preferences, incentives, and capabilities? What are their time horizons?14

These legacies include fundamental cleavages, shared values, and whether a country has a history of stable democracy or has suffered frequent constitutional interruptions. 14 Time horizons are very important determinants of political behavior. Actors with long horizons are much more likely to enter into the intertemporal agreements necessary to sustain effective policies. By contrast, actors with short horizons will tend to maximize short-term political and policy benefits, to the detriment of long-term institutional build-up, and of the credibility and quality of policies. This emphasis on time horizons draws inspiration from an important literature on institutional economics, and its application to politics. See, for instance, Dixit (1996) and references there. 13

Political Institutions, Policymaking,

❚

and

Policy: An Introduction

What are the main arenas in which they interact, and what are the characteristics of those arenas? What is the nature of the political exchanges/transactions they undertake?

Based on their answers to these questions, each of the country chapters provides a detailed discussion of the role played by each of the key players in the policymaking game: the president, political parties, the legislature, the judiciary, the bureaucracy, and in some cases, the governors. In each case, they discuss these players’ roles, incentives, and capabilities. More importantly, perhaps, they bring these actors to life by showing how they engage in political transactions in order to advance their objectives and, more generally, how they play the policymaking game. All these ingredients add up to a thorough portrayal of the workings of the PMP in each case. Having discussed the PMP and the features of public policies, the next section discusses the link between the two. In particular, we highlight the role of cooperation as a feature of the PMP that may help explain the characteristics of public policies, as well as the factors that may foster—or hinder—such cooperation. Policymaking Processes and Policy Outcomes: The Role of Cooperation

We have suggested that important features of public policies depend on the ability of political actors to reach and enforce agreements; that is, to cooperate. For this reason, in addition to other relevant features of the PMP that may be key in shaping policy outcomes in their respective countries, the authors of the country chapters were asked to pay particular attention to a number of factors that may foster—or hinder—cooperation among political actors. Our emphasis on the role of cooperation does not mean that we believe all forms of cooperation are always good. As in the case of collusion among firms in oligopolistic markets, cooperation in policymaking may lead to undesirable outcomes, particularly when some sectors are

Pablo T. Spiller, Ernesto Stein,

and

Mariano Tommasi

excluded from the cooperative agreements and the goals of those who are “in” (for example, retaining power) are different from those of society at large. For cooperation to deliver good outcomes, it needs to be combined with healthy doses of inclusion and political competition. In spite of this important caveat, the type of cooperation discussed in this framework seems to be a key ingredient—perhaps a necessary condition—for several desirable features of good policymaking. Under what conditions is cooperation more likely? Drawing from intuitions from the theory of repeated games,15 it can be argued that cooperative outcomes are more likely if: the immediate benefits from deviating from cooperative behavior are small; there are good “aggregation technologies” so that the number of actors with direct impact on policymaking is small;16 these key actors have long horizons and they interact repeatedly; there are well-institutionalized arenas for political exchange; and there are credible enforcement technologies, such as an independent judiciary or a strong bureaucracy, to which certain public policies can be delegated. The discussion that follows examines each of these conditions in more detail. The elements listed below provide some rough guidance to observe the actual policymaking processes. Clearly, the list is suggestive, not taxonomic. Other important factors may be relevant in characterizing policymaking processes in general, as well as in specific countries. Several historical factors (beyond a specific configuration of political institutions at any moment), as well as cultural, social, and economic configurations, can foster or hinder cooperative political behavior, and affect the PMP in other important ways. Conditions that Foster or Hinder Cooperation

Intraperiod Payoff Structure. If there are large immediate payoffs from deviating from cooperative agreements, cooperation will be hard to sustain.

See Dixit (1996, p. 71), Spiller and Tommasi (2007, pp. 42–46, and references there). Gerring, Thacker, and Moreno (2005) emphasize a related concept: that of “centripetalism.” 15

Political Institutions, Policymaking,

and

Policy: An Introduction

In oligopoly games, for instance, if a firm stands to gain very large shortterm profits by lowering its price (for example, because there are a large number of rival firms from which they can attract customers), it is harder to sustain collusive oligopoly. The same is true in the case of cooperation among political actors. For instance, in the context of the Argentine federal fiscal system, a provinceâ&#x20AC;&#x2122;s individual payoff for deviating from a cooperative agreement (for example, by attempting to get special benefits from the national government, which are paid from a common pool of resources) is quite high; hence the federal fiscal game has noncooperation as its equilibrium outcome (see chapter 3, on Argentina). Number of Political Players. The larger the number of players, the harder it is to cooperate.17 This relates to the previous point in that in many common pool situations, the intraperiod payoff structure is related to the number of players.18 The case of Ecuador, with one of the most fragmented party systems in Latin America, clearly illustrates the difficulties that can be associated with a large number of players in the PMP (see chapter 7). Time Horizons and Temporal Linkages among Key Political Actors. The intertemporal pattern of interactions among specific individuals in formal political positions (such as legislators, governors, and bureaucrats) mat-

Fudenberg and Maskin (1986) and Fudenberg and Tirole (1991, section 5.1.2) make this point, showing that, in repeated games, when the set of feasible payoffs is held constant, increasing the number of players reduces the set of equilibria toward less cooperative ones. 18 Our notion of the number of players is different from the number of veto players in Tsebelis (2002). In Tsebelis, the number of veto players relates (roughly) to the number of actors holding institutional veto positions at a particular point in time. In contrast, we refer to the number of â&#x20AC;&#x153;permanentâ&#x20AC;? players, even if they do not happen to be holding specific veto positions at a particular point in time. For instance, in a country with a stable party system dominated by two major parties that alternate in power, even if one party is out of power at a particular point in time, it is still a player in the intertemporal game. In such a case, we would concur with Tsebelis (2002) and with Cox and McCubbins (2001) in characterizing such a system as decisive. Whether we concur with them in calling that system nonresolute will depend on whether intertemporal cooperation prevents opportunistic short-term policy manipulation. 17

Pablo T. Spiller, Ernesto Stein,

and

Mariano Tommasi

ters for achieving cooperative outcomes. It is not the same to have a legislature in which the same individuals interact repeatedly over extended periods of time as it is to have a legislature where individual legislators are frequently replaced.19 Cooperation is less likely in the latter case, for a variety of reasons. First, longer time horizons tend to lead to lower discount rates, so the benefits of deviating from cooperation today will be smaller relative to the future gains of sustaining it. Second, repeated interaction makes it easier for other political actors to punish those who deviated from cooperation. Third, repeated interaction may also facilitate trust among the different political actors with key roles in the policymaking process. In addition to increasing the scope for cooperation, longer time horizons increase the incentives for political actors to invest in their policymaking capabilities. While there are other factors that affect the nature of the PMP and the features of public policies, it is perhaps not surprising that the countries in Latin America with longer lasting legislators—Uruguay, Chile, and to a lesser extent Brazil—are among the highest ranked in terms of the overall quality of public policies, as measured by the Inter-American Development Bank in a comparative study of the region (IDB 2005).20 Institutionalization of Policymaking Arenas. Arenas that make cooperation easier to enforce can facilitate the complex exchanges required to implement effective public policies. Seminal works on the U.S. Congress disagree on the role that different institutional arrangements (such as the committee system) have in facilitating legislative bargaining, but implicitly agree that somehow things are arranged in a way that facilitates intertemporal cooperation in political exchanges.21 Whether the legislature—as the arena where these transactions take place—is adequately

Beyond the actual tenure of political actors, some historical events, such as coups or frequent episodes of civil unrest, may also increase the discount rate of political actors and limit the scope for cooperation. 20 See IDB (2005, tables 3.6 and 6.1). 21 See Weingast and Marshall (1988); Shepsle and Weingast (1995); Shepsle and Bonchek (1997). 19

Political Institutions, Policymaking,

and

Policy: An Introduction

institutionalized depends on several factors, including legislators’ incentives and capabilities. There are some environments—several in the countries analyzed in this book—in which legislatures are much weaker than in the benchmark U.S. case. In these environments, when political exchanges are undertaken, they tend to take place in settings that are more informal, more uncertain, and more difficult to monitor, observe, and enforce. Delegation. Repeated play encourages self-enforcement of agreements, but certain forms of cooperation can be achieved by alternative institutional means. One alternative is to delegate policy to independent technical agencies. While delegation has its problems, there are instances in which the cost of those problems is smaller than the cost of partisan policymaking. The feasibility of and benefits from such bureaucratic delegation may vary systematically depending on some features of the institutional environment of each country, such as the degree of professionalism of the civil service (Huber and McCarty 2001). The bureaucracy is a key actor for encouraging agreements, especially through its role in putting such agreements into practice over time. A neutral and professional bureaucracy limits the scope for the adoption of opportunistic policies and enhances the trust of actors that commitments made as part of policy agreements will be fulfilled. Several of the country studies look in detail at issues of delegation, which typically introduce important changes into a sector’s policymaking process. The Colombian case (chapter 6), for example, looks at the contrast between policymaking in the fiscal realm—which tends to involve the legislative arena—and in the realm of monetary and exchange rate matters, which were delegated to an autonomous central bank, following the 1991 Constitution.22 Similarly, the Ecuador study (chapter 7) looks at successful delegation in the case of the tax administration, and compares policy features in this sector to those of other fiscal policies,

Delegation to a strong autonomous central bank has also characterized monetary policy in Chile. 22

Pablo T. Spiller, Ernesto Stein,

and

Mariano Tommasi

which are subject to the countryâ&#x20AC;&#x2122;s more general PMP. At the same time, the case of Ecuador, as well as that of Argentina (chapter 3), illustrates how fragile these agreements to delegate policy can be in the context of weak policymaking processes, when delegation becomes politically inconvenient for the incumbent government of the time.23 Availability of Enforcement Technologies. Cooperation is easier to achieve if there is good third-party enforcement. The presence and characteristics of an impartial referee and enforcer of political agreements, such as an independent judiciary, vary from country to country. As a result, the degree of enforcement of political cooperationâ&#x20AC;&#x201D;and the incentives to engage in cooperationâ&#x20AC;&#x201D;vary as well. An Illustration

So far, we have discussed a number of factors that may lead to enhanced cooperation among the political actors that participate in the PMP. But we have not yet discussed how these factors (and cooperation more generally) can affect the features of public policies. While for the most part we leave this discussion to the country chapters themselves, it is worthwhile at this point to provide an example to see how some of these factors may come into play in explaining, for instance, policy stability. Consider once more the hypothetical example of a country with two similarly sized parties that alternate in power. If these parties are highly polarized and their policy preferences differ very significantly, the gains from deviating during each period will be large. Unless there are other factors that lead to intertemporal cooperation, it is likely that each party will try to adopt its preferred policies when it is in office, leading to unstable policies. If in contrast their preferences are more closely aligned, the immediate gains from deviating from cooperation

Costa Rica, a country not covered in this study, presents an interesting case, in which many areas of policy have been delegated to autonomous institutions. See Lehoucq (2007). 23

Political Institutions, Policymaking,

and

Policy: An Introduction

are smaller, and thus cooperation—and policy stability—will be more likely. While the break-up of cooperation may yield immediate gains for incumbents, it also produces long-term costs for them, since deviating today means that policies will be further from their preferences when the opponent is in office, and that the costs associated with frequently switching policies will have to be borne, as well. Whether the immediate benefits compensate for the long-term costs will depend on the rate at which political actors discount the future. If discount rates are very high, cooperation is unlikely to arise. If discount rates are low, political actors will be more likely to agree on an intermediate policy—a “Politica de Estado”—that can be sustained over time. Discount rates, in turn, are determined in part by the length of the actor’s horizons, including such factors as the tenure of presidents, legislators, and party leaders, and the degree to which political parties are long-lived. If the immediate gains from implementing their preferred policies (rather than a compromise) are fairly low, and so are their discount rates, political actors will have incentives to behave cooperatively. Still, they will need some mechanism by which they can reach the necessary agreements and broker the necessary deals, which may involve compensation in other policy areas or across time. In the absence of adequately institutionalized arenas where these deals can be brokered and upheld over time, it will be difficult for the parties to reach these deals. In some cases, the immediate benefits of deviating from cooperation—or the discount rates—may be large enough that cooperation will not be self-enforcing. While cooperation may lead to better results for everyone ex ante (for example, before the results of the elections are known), once an incumbent is in office, he or she may have incentives to deviate. In such cases, political actors may want to agree to an intermediate policy beforehand, and put in place mechanisms that will ensure that the deal is upheld. Delegating policymaking to a competent autonomous bureaucracy is one way to do this—provided the delegation is credible. An independent judiciary may help make such delegation or other forms of commitment more credible, by forcing political actors to stand by them.

Pablo T. Spiller, Ernesto Stein,

and

Mariano Tommasi

Political Institutions and the Policymaking Process

The discussion so far has focused on public policies, the policymaking process, and on the connection between them. As figure 1.1 illustrates, the PMP is in turn determined by each country’s political institutions, at least in part. But which specific institutions matter? And what are the channels through which they matter? One of the main lessons from this comparative project is that there are a myriad of institutional dimensions that can have an impact on the workings of a country’s policymaking process, and that the impact of each of those dimensions tends to be contingent on the configuration of other institutional dimensions. Chapter 2 by Carlos Scartascini surveys the rich literature that addresses the way in which political institutions (such as constitutional norms and various aspects of electoral rules) shape the roles, incentives, and modes of engagement of the main players that participate in the policymaking process. Chapter 2 is organized actor by actor, looking in turn at presidents, cabinets, the legislature, the judiciary, subnational actors, political parties, and bureaucracies. Thus the chapter tends to highlight the rules and practices that have an impact on the way each of these actors plays the policymaking game, rather than highlighting the interaction among different policy players and institutional dimensions. While this background is of course critical, one of the main points that we emphasized in the guiding framework and that comes across very clearly in the country studies is the importance of interactions, of configurations of institutions, and of the need for a systematic approach to the complex subject of policymaking. Looking at the eight country studies in this book, one gets the clear message that the impact of political institutions on the PMP is configural. Each country’s policymaking process is the result of a configuration of political rules and practices, shaping the preferences, incentives, and constraints of policymaking actors. A focus on a few institutional characteristics—such as whether the country has a presidential or parliamentary system, whether the electoral rules are of the plurality or proportional representation variety, or whether the president has

Political Institutions, Policymaking,

and

Policy: An Introduction

more or less partisan or legislative powersâ&#x20AC;&#x201D;will provide only a very fragmented and unsatisfactory understanding of these processes. In order to understand them fully, the institutional set-up needs to be looked upon as a system.24 The point can be seen most clearly if the institutional factor one focuses upon is the very â&#x20AC;&#x153;broadâ&#x20AC;? category of presidentialism/parliamentarism. All eight cases studied in this book are presidential, yet policymaking processes, as well as policy outcomes, are very different across countries.25 But even countries that look similar in several of the relevant dimensions that have been emphasized in the literature (and that are surveyed in chapter 2) may have very different PMPs and policy outcomes. Consider the cases of Brazil and Ecuador. In addition to being presidential systems, both these countries share an extreme degree of political party fragmentation. In fact, they have the most fragmented party systems in Latin America, which translates into the lowest share of seats controlled by the party of the president (see Jones 2005; IDB 2005). In both countries, legislators are elected under proportional representation, with open lists. Thus in both, legislators have incentives to deliver benefits to their communities. In both, presidents have strong legislative powers to compensate for the weak partisan powers. Yet in spite of all these similarities, these countries are very different in terms of their policymaking processes, as well as their political and policy outcomes. While Brazil has embarked on a path to good governance and presents solid public policies in many areas, Ecuador is characterized by political instability and policy ineffectiveness. A partial look at a few institutional dimensions, as important as these may be, is clearly insufficient

Other authors have emphasized the importance of configurations and interactions among different institutional dimensions: see Liphjart (1991); Tsebelis (1995, 2002); Haggard and McCubbins (2001); and Fukuyama (2006). We believe that this is the first book to study this point systematically in a number of integral and detailed country studies. 23 Other studies taking advantage of variations within presidential regimes include Mainwaring and Scully (1995a); Mainwaring and Shugart (1997a); Carey and Shugart (1998b); Haggard and McCubbins (2001); Morgenstern and Nacif (2002); and Payne and others (2002). 24

Pablo T. Spiller, Ernesto Stein,

and

Mariano Tommasi

to explain the characteristics of the policymaking process and policy outcomes. Obviously, there are some other institutional dimensions that matter, and the interaction between these different dimensions matters as well. Sometimes seemingly smaller details of the institutional setting, such as the discretion of the president to allocate budget resources that generate local benefits (possible in Brazil, not in Ecuador), can play an important role (see chapters 4 and 7, as well as IDB 2005). The chapters in this volume present various other instances in which institutional rules usually associated with some outcomes â&#x20AC;&#x153;in partial equilibriumâ&#x20AC;? are not associated with those outcomes in specific country cases, since the configuration of political incentives make them operate differently. For instance, the president of Chile is constitutionally the most powerful president in the region. Yet the political dynamics of the country, as well as a wealth of complementary institutions, have led the presidents since the return to democracy to exercise that power in a careful and consensual manner, following a practice of negotiation, compromise, and consensus building. Such policies have fostered the development of credible and stable policies, while allowing presidents to (gradually) imprint their own priorities on the countryâ&#x20AC;&#x2122;s agenda (see chapter 5 on Chile; Weyland 1999). At the other extreme, the president of Mexico has traditionally been very weak in terms of his constitutional powers. Yet during the PRI era, he dominated the policymaking process, since the weak constitutional powers did not really matter much in the context of single-party rule.26 The Country Studies

The comparative research project reflected in this book steered a middle course between imposing a common methodological framework and giving the authors of the country studies the freedom to enrich the analysis of their cases both methodologically and empirically. The authors of the subsequent chapters have done a terrific job of judiciously applying the As chapter 8 on Mexico discusses, with the arrival of divided government, the constitutional weakness of Mexican presidents has come to the fore. 26

Political Institutions, Policymaking,

and

Policy: An Introduction

general methodological guidelines summarized in this introduction, while giving the idiosyncrasies of each case the weight they deserve in order to capture the essence of each country’s policymaking process. Some of the chapters have implicitly or explicitly broadened the basic guiding framework in some dimensions important to describe their cases better. Some country studies (such as Brazil and Ecuador) have framed the description of the policymaking game from the perspective of the ability of the president to pass his or her agenda through congress, and the ease or difficulty of implementing exchanges along alternative dimensions, such as making policy compromises, filling cabinet positions, and distributing resources to specific constituencies. Other studies have focused in more detail on the specific distribution of preferences represented in the PMP of their countries, enabling the simultaneous explanation of both general characteristics of and specific winners and losers from policy in different areas (Chile) or time periods (Mexico). In order to characterize the policymaking process of each country, authors have combined firsthand knowledge of key policy actors and their interactions with secondary sources—and, in many cases, new empirical work. Each chapter contains a nuanced mix of quantitative and qualitative research techniques. For example, in chapter 4 on Brazil, Lee J. Alston, Marcus André Melo, Bernardo Mueller, and Carlos Pereira discuss several episodes in which the supreme court ruled against the executive on issues of vital importance, in order to back up their claim that the Brazilian supreme court has been relatively independent. At the same time, they present novel empirical evidence about the connection between votes for the government in congress and the appropriation by the executive of legislators’ budget amendments, providing support for their argument about the centrality of the exchange of pork for political support in the midst of a highly fragmented party system. Similarly, the chapter on Ecuador provides a detailed account of the mechanics and the inducements with which the government attempts (not always successfully) to form and maintain coalitions, also in the context of a very fragmented party system that leads to considerable political and policy instability. In the chapters on Colombia and Paraguay, the au-

Pablo T. Spiller, Ernesto Stein,

and

Mariano Tommasi

thors illustrate their claims about the changing role of the legislature in the policymaking process by carrying out empirical analysis of legislative activity in the different periods they consider. This novel econometric analysis was enriched by the fact that in the Colombian case, two of the authors are former ministers with extensive experience in working with congress, and that the Paraguay team has drawn on background work based on extensive access to interviews with key actors. A number of the studies exploit variations within a particular country, either across policy issues or across time. Variation of policy characteristics across issues is explained in the Brazilian case by arguing that those policies most important in the presidential agenda will be stable and adaptable; that many other policies will be volatile, since the resources devoted to them will depend on the budget situation; and that some policies with strong interests will be hard-wired into the constitution precisely to escape that volatility. The chapter on Chile argues that policies will have all the desirable properties (stability, adaptability, good implementation) in those cases where preferences are not too far apart (such as macroeconomic policies), but â&#x20AC;&#x153;veto-playerâ&#x20AC;? effects will dominate when preferences are very polarized (such as moral issues). Several of the countries studied in this volume experienced major changes in their political institutions during the period under study. Colombia, for example, introduced a new constitution in 1991, which led to a more active role for congress, the judiciary, and subnational authorities, thus imposing increased constraints on the authority of the executive. Venezuela introduced elections for governors in 1989, which substantially changed the political and party dynamics, and led to important changes in policy outcomes. The 1999 Constitution introduced during the ChĂĄvez administration brought further changes that concentrated power in the hands of the president and considerably altered the nature of policymaking in this country once again. In Mexico and Paraguay, the process of policymaking changed significantly with the end of the PRI hegemonic period and the Stroessner dictatorship, respectively. The authors of these chapters take advantage of this variation across time, and attempt to pin down the impact of these changes on policy outcomes.

Political Institutions, Policymaking,

and

Policy: An Introduction

While the country studies that exploit variation of political institutions across time provide some insights into the reasons behind the changes in the relevant rules of the game, we should stress that the focus of this book is not on the dynamics of institutional change, but rather on the effects of political institutions on policymaking and on policies. That is, the country chapters by and large take political rules as given and focus on their implications.27 We are aware, of course, that institutions are themselves endogenous and that a more complete understanding would also require an analysis of institutional change and of the endogenous determination of political rules. Yet from a rational choice perspective, any theory attempting to explain how institutions change should also understand, as a building block, the consequences associated with the different institutional settings (Diermeier and Khrebiel 2003). Since one project cannot cover all bases at the same time, we leave the study of the evolution of institutions for a later stage of our research agenda. Parting Thoughts: The Challenges of Institutional Reform

To close this introduction, we come back to one of the original motivations behind this research agenda, and provide some brief reflections about institutional reform. There is a new wave of thinking about economic policy in developing countries that argues against universally applicable policy prescriptions being pushed in all countries, regardless of It is interesting to notice that some of the country studies suggest explanations of institutional change that draw from the framework of intertemporal cooperation of this project. For instance, in Venezuela, a cooperative power-sharing agreement between the two dominant parties that began in 1958 with the Pacto de Punto Fijo, started to break down during the Pérez administration (1974–78) in the context of the first oil crisis. As argued by Francisco Monaldi, Rosa Amelia González, Richard Obuchi, and Michael Penfold (chapter 10 in this volume), the huge windfall of oil resources increased the informal powers of the president, and gave way to presidential unilateralism, reflected in a dramatic expansion of the use of presidential decrees. This breakup of cooperation at a time of a windfall of oil money to the country is consistent with the prediction of the theory of repeated games that the benefits of deviating from cooperation are larger when short-term resources are greater than those expected in the future. Similarly, the case study of Paraguay relates the unraveling of the Stroessner-era equilibrium to the “last-period problem” during the late Stronismo. 27

Pablo T. Spiller, Ernesto Stein,

and

Mariano Tommasi

the circumstance, time, and place where they are applied.28 The previous discussion regarding the complexities involved in the link between political institutions and policymaking processes, as well as the important role of interactions between different institutional dimensions, suggests that one should adopt a similar logic when thinking about institutional reform.29 The merits of potential changes in political and institutional rules must be considered carefully, with an understanding of how these rules fit within the broader institutional configuration. Broad generalizations about the merits of different political regimes, electoral systems, or constitutional adjudication of powers among branches are not very useful. Partial equilibrium views that stress the importance of a single institutional dimension may lead to misguided institutional and policy reforms. Understanding the overall workings of the political process and of the policymaking process in each specific country, with its specific historical trajectory, is a crucial prerequisite for developing appropriate policy reform proposals, as well as institutional reform proposals. In addition, we believe that one needs to be very cautious in attempting a technocratic solution to what are essentially political problems. The incentives of professional politicians such as presidents, legislators, and party leaders can impose important constraints on the type of reform that is feasible. Adopting the best civil service law in the world will not work if patronage involving positions in the bureaucracy remains an important currency used by politicians to reward their partisan base. Attempts to improve the policymaking capabilities of congress will not work if legislators themselves do not have incentives to develop such capabilities. This does not mean that such reforms cannot take place. It just means that success in this area may require creative ways of advancing reform that take the incentives of the key players into account.

See, for instance, Evans (2001); Lindauer and Pritchett (2002); Hausmann, Rodrik, and Velasco (2005); and IDB (2005). 29 For a similar view, see North (1994); Pistor (2000); Evans (2004); and Eggertsson (2005). 28

2 Whoâ&#x20AC;&#x2122;s Who in the PMP: An Overview of Actors, Incentives, and the Roles They Play Carlos Scartascini *

Introduction

Chapter 1 presents a framework linking political institutions, policymaking processes, and policy outcomes. Within that framework, public policies are seen as the outcome of the interaction among a variety of political actors. These actors, each with its own preferences and incentivesâ&#x20AC;&#x201D;and within the constraints of the rules that frame their engagementâ&#x20AC;&#x201D;meet in different arenas to define public policies. This chapter looks at a number of those key actors, institutions, and arenas, with the aim of examining the roles, incentives, and capabilities of each of the actors in the policymaking process, and thus their impact on the features of public policies. This chapter, which draws from an extensive literature in political science and political economy, does not pretend to be a complete survey; rather it highlights those institutions and actors that are important within the context of the framework developed in chapter 1, and that tend to have a key role in most of the countries studied in this volume.1 While

This chapter draws on Scartascini and Olivera (2003) and has benefited greatly from comments by Ernesto Stein, Mariano Tommasi, and two anonymous referees. Martin Ardanaz provided superb research assistance. The usual disclaimer applies. 1 Other sources that should be considered for a complementary, more comprehensive, and substantial overview include Weaver and Rockman (1993); Mueller (1996a, 1996b, 2003); Carey (2000); Persson and Tabellini (2000, 2003); Haggard and McCubbins (2001); Payne and others (2002); Tsebelis (2002); and IDB (2005). *

Carlos Scartascini

the framework emphasizes the interaction among actors, this chapter examines them individually, for the most part. The richness and complexity of interactions are more naturally addressed within the realm of the specific country cases, and therefore are mostly left to the country chapters that follow. Each section of the chapter analyzes specific actors in the policymaking process, as well as the political institutions shaping their roles and incentives. For the most part, the order in which we present the institutional actors follows the traditional layout of institutions presented in the constitution of a country. Usually, democratic constitutions assign the role of policymaking to three separate but related branches—the executive (and cabinet), the legislature, and the judiciary—establishing the prerogatives, functions, and scope of these institutions. Additionally, there is sometimes a vertical dimension of institutions (federalism) that regulates the relationships between the central and subnational governments. Finally, the chapter analyzes the role of other actors with formally ascribed roles in the policymaking process, such as political parties and the bureaucracy. The Executive Branch: Presidents and Cabinets

Countries can organize their executive branch along a “continuum” between two polar cases: presidential and parliamentary systems.2 The choice of political regime between presidential and parliamentary systems can have important consequences for policymaking because it has an influence on the number and stability of the agents in charge of policymaking, the arenas where exchanges take place, the type of political exchanges that can take place between the executive and the

Lijphart (1999) identifies three basic differences between presidential and parliamentary systems. First, the executive in a parliamentary system is responsible to the legislature and can be dismissed from office by a legislative vote of no confidence or censure. In a presidential system, the head of government is elected for a constitutionally prescribed period. Second, prime ministers are selected by the legislature, while presidents are popularly elected. Third, parliamentary systems have collective or collegial executives, whereas presidential systems have single person, noncollegial executives. 2

Who’s Who

in the

PMP

legislature, and some of the bargaining prerogatives of each of the actors. Latin American countries have traditionally opted for presidential regimes, instead of the parliamentary systems that are more popular in Europe. Even though this study focuses on presidential systems, it is important to characterize both systems, insofar as this discussion helps introduce the literature and stylized facts on stability of governments and policies. In fact, some critics argue that some of the problems with policymaking in Latin America have their origin in the region’s choice of political regime. The stylized facts indicate that parliamentary systems tend to be less stable than presidential systems because political government leaders tend to change more frequently. However, when changes occur in parliamentary systems, they are usually smooth and do not involve a complete reshuffling of the government; some of the policymakers remain in place, along with their policies. On the other hand, changes in presidential systems are more dramatic when they occur because they entail either a democratic breakdown or a major reshuffle of the government. In that context, Linz (1990, 1994) has characterized presidential systems as “rigid” and parliamentary systems as “flexible.” In his work, flexibility is to be preferred to rigidity, especially because flexibility is risk-minimizing (for example, crises in parliamentary systems would be government crises, and not regime crises). Thus the rigidity of presidentialism, crystallized in situations where presidents lack a majority of seats in legislatures, was thought to be one of the main determinants of the breakdown of democratic regimes in Latin America (Linz and Valenzuela 1994). This argument regarding the relationship between minority governments and regime survival has been tested empirically by recent scholarship. On the one hand, based on data for all presidential democracies that existed between 1946 and 1996, Cheibub and collaborators show that minority presidents, minority governments, and deadlock situations do not affect the survival of democracies (Cheibub 2002; Cheibub and Limongi 2002; Cheibub, Przeworksi, and Saiegh 2004). On the other hand, based on a much smaller sample of Latin American governments

Carlos Scartascini

between 1978 and 2005, Chasquetti (2004) and IDB (2005) note that in this period, a number of minority governments (defined as situations where governments control less than 45 percent of legislative seats) suffered constitutional interruptions (situations where either the president or congress does not finish the terms for which they were elected). For this selected group of countries, IDB (2005) finds that minority governments (whether single-party or coalitional) were five times more likely to suffer constitutional interruptions than governments with a majority or near majority of seats. Considering the fact that the modal type of party system in Latin America is a multiparty one, the ability of governments to form and maintain majority coalitions should be considered an important factor that may affect political stability (or the lack thereof) in the region. In the case of Ecuador, a country where presidents usually have a minority of the seats in the legislature (lately around 25 percent) and where coalitions tend to be unstable, no president has finished his four-year constitutional mandate since 1996. Presidents

The regime type adopted in Latin America makes presidents key players in the policymaking game. Therefore, it is important to understand their incentives and the factors affecting presidential behavior. Even though their personal qualities, ideology, and historical and cultural factors can shape the way presidents govern, the institutions that determine the way they are elected and their power to affect policy decisions tend to be at least as important in explaining their incentives and behavior. In some institutional contexts, presidents tend to be mostly interested in the public good and design their policies taking into account broad interests in society. In other institutional contexts, presidents have â&#x20AC;&#x153;mixedâ&#x20AC;? incentives, and are driven by personal and political goals that may interfere with the goal of serving the general public interest. While differences in incentives are explained mostly by electoral rules, the capacity to transform policies depends on the powers bestowed on the presidents, which are discussed next.

Who’s Who

in the

PMP

Presidential Powers. The powers of the president determine the strategic actions the president may take, and the type of transactions he or she may engage in with the legislature and his or her political allies and opponents. Presidential powers can be classified as either constitutional powers or partisan powers (Shugart and Carey 1992; Mainwaring and Shugart 1997a). Constitutional powers contribute to frame the relationship between the executive and the legislative. As the constitutional powers of the presidency increase, other things being equal, so does the president’s discretion to introduce changes to the status quo. Higher constitutional powers make it easier for the president to take decisive action whenever it is necessary (such as adapting economic policy to shocks), but they also make it easier to change policies for political convenience, leading to a potential volatility of policies (such as changing policies before an election for political gain or discarding the policies of the previous administration and “starting all over” after taking office). Constitutional powers can be divided into legislative and nonlegislative powers. Legislative powers include the package veto, the partial (line item) veto, the power to issue decrees or declare a bill “urgent,”3 the exclusive initiative of legislation, budgetary powers, and the power to call a plebiscite or referendum. These powers can be further divided into proactive and reactive, according to whether the president can influence the adoption of policies that represent a change in the status quo or whether the president can stop or delay the implementation of policies that could modify the status quo (Carey and Shugart 1998b). If legislative presidential powers are important, policies will be closer to the preferences of the executive branch. For example, the president could obtain policies closer to his or her preferences by threatening to veto or actually vetoing the legislature, by issuing decrees and thus bypassing the legislature, and by using the prerogative to consult the citizenry through referendums to

Decree is the authority of the executive to establish law without prior consent of the assembly (Carey and Shugart 1998b). This may include executive policy initiatives that eventually require legislative ratification. Urgency bills are proposals issued by the president and become law unless the legislature acts to reject them within a specified time period. 3

Carlos Scartascini

bypass the opposition of the legislature for those policies for which the president can ensure popular support (Mueller 1996a, 1996b).4 The president’s nonlegislative powers include the power to nominate, appoint, and dismiss government officials. The rules for cabinet formation and cabinet dismissal affect not only the power of the president within the executive branch but also the relationship between the executive and legislative branches. If the rules provide the legislature with power over the appointment and dismissal of cabinet ministers, legislators could use those prerogatives as a bargaining mechanism with the executive over certain policies of their interest.5 A related bargaining chip of the legislature is the mechanism for impeachment of the president, which could affect the degree of vulnerability of the president vis-à-vis congress (Pérez Liñán 2006). Constitutional rules are not the whole story for understanding the bargaining between the branches of government. In fact, these rules can be assessed only in interaction with the capacity of presidents to mobilize support among members of congress (Pérez Liñán 2006). This takes us into the realm of the partisan powers of presidents, a key variable whose importance goes far beyond the impeachment process, and to which we now turn. Partisan powers relate to the degree of support for the president in congress. The standard measures are the size of the president’s legislative contingent and the degree of party discipline (Mainwaring and Shugart 1997a). Dominant parties capable of ruling by themselves (in presidential systems, presidents and legislatures of the same party, especially if the party is a “centralized” one)6 have the easiest time securing legislative

The relevance of the elements of direct democracy (referendum, popular initiative) has been highlighted as one of the resources at the disposal of the executive branch to pass certain policies when the legislature is opposed to them. The literature on popular initiative, based on the tradition of Romer and Rosenthal (1978, 1979, 1982), shows that the agenda-setter has good possibilities of having its policies supported by maximizing support through popular votes, given the preferences of the voters and the status quo (reversionary position). 5 See Shugart and Carey (1992); Amorim Neto (2002); Payne and others (2002); IDB (2005); Martínez Gallardo (2005). 6 Centralized parties are those where national politicians have tight control over the valuable resources needed to further legislative and political careers. 4

Who’s Who

in the

PMP

support for their programs. Coalition governments fare less well, and presidential systems in which the president and legislature are of different parties have the greatest difficulty (Haggard and Webb 2000). Sometimes, presidents can still govern in cases of minority government: they can circumvent potential opposition in the legislature by relying on the legitimacy provided by strong popular support. To a certain extent, the degree of support and legitimacy can be traced back to the electoral system that rules the election of presidents, the electoral system that rules the election of legislators, and the degree of concurrence of their elections. Electoral Rules

Presidents can be directly elected using plurality voting or runoff elections, or they can be indirectly elected, either through an electoral college or by legislative decision (usually as a “second round” mechanism, as in Bolivia). The method of election of the president is particularly relevant, as it affects the degree of popular support of the president and determines whether forming a coalition is necessary for obtaining the presidency. Electoral systems that ensure a high share of the vote for the winning candidate (particularly in a first round) tend to increase the legitimacy and reduce the need for coalitions. As discussed in the next section, the electoral rules of the legislature also affect the extent to which coalitions are needed to govern by affecting the degree of fragmentation of congress—and thus the probability that the president could get a majority in congress. Having to rely on coalitions to govern usually creates restrictions for the president in the bargaining process, as the chapters on Brazil, Chile, and Ecuador show. In addition to the formulas that transform votes into seats, other institutional aspects of electoral rules are important, including the concurrence of elections and the use of midterm elections. When elections are concurrent for the two branches, the president’s party is likely to receive more votes, and fewer parties are likely to receive significant shares of the vote in legislative elections. This reduces fragmentation and increases the chance that the president can win strong legislative support (Jones 1997; Shugart

Carlos Scartascini

1995). Midterm elections, whether in systems with or without coterminous cycles, can contribute to difficulties in governing, mainly if opposition majorities result (Shugart 1995). These elections can weaken the policymaking effectiveness of the executive by altering the balance of partisan power mid-way through the term, and by shifting congressional attention from the policy agenda to electoral strategizing and campaigning. In addition to explaining the president’s relationship with the legislature, it is also relevant to explore the determinants of the president’s relationship with the party. One of those determinants is the nomination procedure of presidential candidates. Some of the systems increase the allegiance of presidents to the party, while other systems tend to favor the appearance of “extra-party” candidates (Morgenstern and Siavelis 2004). For example, when the selection of presidential candidates is centralized (controlled by party elites) and high barriers for independent candidacies exist, it will be more likely for party insiders to become presidential candidates.7 By contrast, decentralized recruitment and low barriers to independent candidacies may encourage the appearance of candidates known as “freewheeling independents.” These candidates have no long-term identification with a party and typically use parties as mere electoral vehicles to reach the presidency. As a consequence, elected presidents will be less constrained by party ties, but they may be unable to build legislative coalitions (Morgenstern and Siavelis 2004). In this case, policies could become more unstable (because the policies pursued by the president are the president’s own and do not follow a historic party stance) and less adaptable (because it will be harder—usually more costly—to respond to shocks). The incentives behind some of the actions of the president may also be explained by the tenure of the presidency, given term limits and re‑ election constraints, because they explain their decisionmaking horizon. If presidents can serve consecutive terms and there is the possibility that presidents can be reelected, the policies pursued by the president will usually be influenced by the reelection campaign. If there is no possibility of reelection, the policies presidents will try to enact will be a mixture of Party insiders are candidates who emerge from long-standing, institutionalized, and programmatic parties, and who have held positions in the party before becoming candidates. 7

Who’s Who

in the

PMP

trying to help the candidate of the party (mainly if former presidents tend to keep some role in party politics) and trying to influence the set of policies that the next president will have at his or her disposal (Carey 2003).8 Therefore, a candidate’s potential post-tenure career path could be very important in defining the policies he or she pursues while in office. Cabinets

The cabinet—even if it is not always an initiator of policies (which it is in many countries)—is usually a major player for the attainment of the government objectives because it is in charge of the actual implementation. Additionally, cabinets usually serve as a mechanism to cement coalitions. Given certain constitutional mandates that regulate the formation of the cabinet—which are usually determined by the type of political regime—the electoral system and the party system could have an impact on the number of ministers, the issues under their domain, the way in which they are appointed, their capacity for coordination, their responsibilities, and their political allegiance. More importantly, the electoral system and the party system could affect the duration of the ministers, their level of specialization and skills for the task at hand, and their mobility (Blondel 1985). Among the characteristics of cabinets relevant for policymaking, two are the most salient: the process of cabinet formation and the stability of cabinets (Martínez Gallardo 2005). The process of cabinet formation affects the identity of the cabinet and the allegiance of its members to the president. The decisions on who to name to those (sometimes) key positions depend in part on the electoral system for the presidency. Electoral systems that ensure a majority for the president’s party do not generate the need for the formation of coalitions and will usually result in compact cabinets (that is, cabinets made up entirely of members from the president’s party).9 On the other hand, when Offering past presidents a position in the party could be a mechanism for aligning their incentives when leaving office with those of the party. 9 A compact cabinet would have fewer conflicts over policy. For example, Bawn and Rosenbluth (2006) find that the fewer the number of parties in the government coalition, the lower are public expenditures. 8

Carlos Scartascini

coalitions are needed, they can be cemented through positions in the cabinet, generating multiparty cabinets. The formation of the cabinet becomes more relevant the weaker the partisan powers of the president, or the more fragmented the party system. For example, in the case of Bolivia, there is a strong correlation among the number of ministries, the number of cabinet positions offered to coalition members, and the share of the votes obtained by the president in the general election (IDB 2005). In cases of government coalitions, the degree of cabinet coalescence (that is, the extent to which cabinet posts reflect the distribution of seats held by the parties joining the executive in the legislature) may affect legislative voting behavior, and thus the ability of the president to pass his agenda. For example, Amorim Neto (2002) finds that cabinets in Brazil that display a high correspondence between cabinet portfolios and legislative seats held by the parties joining the executive foster coalition discipline, as they generate incentives for the parties to support executive initiatives in the legislature. While President Cardozo’s cabinet maintained that correspondence, President Lula’s cabinet was more partisan. According to Pereira and Power (2005), this is one of the reasons why his government had to cement the coalition through other means in congress. Cabinet stability affects the stability and efficiency of policymaking, and thus of public policies.10 One source of cabinet instability is frequent changes in government. In Latin America, it is common for each incoming president to change not only the people in charge of the ministries but also the overall structure of government by creating new ministries, eliminating others, changing their names, functions, and scope, and the like.11 High cabinet instability ultimately leads to high policy instability, low accumulation of expertise, and thus low quality and less adaptable policies. When rotation is high, ministers have no time and incentives to accumulate expertise. These “negative” incentives usually trickle down to the lower levels of the bureaucracy. For example, Amorim Neto and Borsani (2004) find that cabinet stability (low ministerial turnover) is conducive to fiscal policy stability: that is, the ability to control spending and attain fiscal balance. 11 In Latin America, it is common to see “failed innovations”: ministries that are created and abolished a few years later (Blondel 1982). 10

Who’s Who

in the

PMP

The Legislative Branch

Two roles have traditionally been conferred on the legislature: to legislate and to restrain the executive branch. In terms of the framework of this study, the legislature, since its inception, has been considered an institution that would help reduce the volatility of polices and better represent the preferences of the populace. The legislature’s effectiveness in doing so depends on the institutions of the executive just described and on the institutions of the legislature. Legislative institutions (broadly defined) can have important consequences for how policies are decided and implemented and the possibility of engaging in intertemporal cooperation. In terms of the approach followed in this book, those institutions could have an impact on the number of relevant political actors and their relative duration (and discount factor), tilt the results of the policymaking process in favor of some geographic or demographic minority, affect the “quality” of the representatives, and affect the arenas where decisions are made, among other possibilities. Among the legislative institutions that could affect the behavior of legislators, this discussion highlights two factors: electoral rules, the rules of access to the legislature, such as the electoral rules and party nomination procedures; and legislative structure, the rules that organize the workings of the legislature, such as whether the legislature is divided in two chambers, and the roles and prerogatives of committees.12 Electoral Rules

As is the case with the executive branch, the method of election of representatives is a very important determinant of the number of relevant policymakers and the incentives of policymakers.13 One way the traditional

These two groups are interdependent because the rules and workings of congress tend to be an endogenous response to the impact of the electoral rules on party discipline and organization. The study of the inner workings of congress and its interactions with the executive and other actors is one of the main focuses in the country cases covered in this volume. 13 This discussion does not follow very strict criteria for classifying electoral systems. Using somewhat stricter criteria, Katz (1997) classifies electoral systems according to: 12

Carlos Scartascini

literature has analyzed the impact of electoral rules on policymaking has been primarily through their impact on representativeness, effectiveness, and participation.14 An electoral system that fosters these features would basically ensure that a strong connection exists between citizens and representatives, that citizensâ&#x20AC;&#x2122; preferences are represented in the legislature according to their weight in society, and that legislators can work those preferences into policy effectively. The first characteristic of the method of election is whether representatives to congress are directly elected, indirectly elected, or appointed. These options influence the number of relevant players. Because representatives have as one of their main objectives maximizing the returns on their political careers, they will usually try to satisfy those constituencies that provide them with the greatest benefit. This choice can differ according to the method of election. For example, if representatives are appointed by the executive branch, they are usually not potential veto players for executive branch initiatives; this reduces the number of relevant actors. If representatives are selected by the subnational legislatures, they could become highly relevant players when dealing with some issues, such as intergovernmental transfers; this increases the number of players, at least with regards to these issues. In Latin America, indirect election and appointment have been gradually phased out in favor of direct election. While this reform alone could help to align the preferences of the representative with those of the citizenry, differences in the system used for the direct election of representatives play a role as well. The most common differences are those translation of votes into seats (electoral formula and constituencies); nature of choice (object of choice, type and number of choices); access to the ballot box (suffrage, registration of voters, ease of voting); and control of candidacy (qualification and nomination, campaign activity, public subsidy). 14 An electoral system that is optimally representative is one in which political groups obtain legislative seats in nearly exact proportion to their share of the vote. An electoral system fosters effectiveness if it produces sufficient concentration of power in the legislature to make it possible for diverse societal preferences to be aggregated and resolved into acts of government (Payne and others 2002). Finally, participation refers to how the form of voting affects the strength of the connection between the constituent and his or her representative. This is also called the agency dimension, and it could affect voter turnout (Grofman and Reynolds 2001).

Whoâ&#x20AC;&#x2122;s Who

in the

PMP

with respect to the electoral formula, the ballot structure, and/or the magnitude of the electoral districts.15 In some countries, each district is represented by a single legislator (single-member districts); in other countries, each district is represented by several legislators (multimember districts). In single-member districts, the seat is allocated to the candidate who had the largest number of votes (either in the first ballot or after a runoff election). In the case of multimember districts, candidates are usually part of a party list, and seats are allocated according to one of a variety of different formulas that relate the percentage of votes that the candidates or group of candidates received in the election to the percentage of seats. Ideally, candidates who receive a given share of votes would receive the same share of seats. However, the electoral formula plus the district size (how many legislators are elected from each district) and the thresholds of representation (the minimum percentage of votes that a party must reach to win representation) could introduce a wedge between the share of votes and the share of seats, increasing the degree of disproportionality, and thus affecting the degree of influence of different actors in society.16 The electoral rules, particularly the district magnitude, could also affect whether those who try to enter the legislature do so through an existing party or a new party (because of strategic voting).17 Effectively, The electoral formula is the method by which vote totals are translated into claims upon seats. The main classes of electoral formula in democracies are plurality rule and proportional representation. The ballot structure consists of the number of votes each voter is entitled to cast; whether voters are allowed to abstain from using some of their votes, when they have more than one, or must cast them all; and whether voters can cumulate their votes or not. Finally, the magnitude of the electoral district refers to the number of seats to be filled by the voters of that district. 16 The disproportionality for each party in a particular election is simply the difference between its vote share and its seat share. Disproportionality is usually larger in countries that use majority or plurality voting than in countries that use proportional representation systems. Among those that elect their legislators using proportionality, disproportionality would be larger the smaller the size of the district; the higher the threshold of representation; and if countries have presidential systems with concurrent elections. 17 Strategic voting refers to a type of behavior induced by certain electoral rules in which voters choose not to vote their first order candidates, in order to prevent the least preferred candidates from winning (Shepsle and Bonchek 1997). 15

Carlos Scartascini

the number of effective parties (and potential players) is usually higher as the system becomes more proportional. In Latin America, the effective number of parties is close to two in Chile (with a district size of two) and almost eight in Brazil, where the average district is close to twenty (IDB 2005).18 The relationship between the number of parties and the number of relevant players in the legislature depends on the degree of allegiance of candidates to parties.19 If party discipline is high, the number of effective political parties would proxy the number of agents with power over decision making in the chamber. Party discipline is also affected by the electoral system because it shapes the incentives of those who must compete, particularly regarding their allegiance to the party bosses or to the constituency that elects them. Under a regime of multimember districts (large size districts) and proportional representation, party leaders have a higher number of “carrots and sticks” (particularly under closed lists); thus politicians respond to the party leadership’s platform to increase their chances of nomination (Gallagher, Laver, and Mair 1992).20 Under a regime of single-member districts and plurality rule (and to a lesser extent under proportional representation with small district size), politicians can usually act as political entrepreneurs who respond mostly to their local constituency to secure nomination and reelection. The degree of fragmentation of a legislature is usually measured one of two ways: through Rae’s fragmentation index (Rae 1967), which indicates the probability that two randomly chosen legislators would belong to a different party; or through the effective number of parties (Laakso and Taagapera 1979). The major advantage of the effective number of parties is that it can be visualized more easily than the Rae index (it approximates the equivalent number of parties of equal size for a given fractionalization). 19 One way to look at the incentives of candidates to conform to citizens or to their party is by looking at the degree of particularism: that is, the incentives to cultivate a personal or individual vote versus a party or collective vote (see Carey and Shugart 1995; Wallack and others 2003; Johnson and Wallack 2006). 20 Rasch (1999) considers that holding equal other institutions, electoral systems have an impact on party discipline through three different channels: district magnitude, ballot, and decision rules or electoral formulas. Basically, parties represented in the legislature would be more disciplined if there is a large average number of seats per district, if the placement on party lists is centrally controlled or is controlled by the party branch of large regions, and if formulas are more proportional (there is a low threshold for election). 18

Whoâ&#x20AC;&#x2122;s Who

in the

PMP

In addition to the size of the district, in the case of multimember districts, another important consideration is whether legislators are elected from closed or open lists (ballot structure). The ballot structure has important implications, as it could affect electoral strategies, the degree of party discipline, and the link between voters and representatives (Carey and Shugart 1995). Assuming that party labels are meaningful, closed list systems provide party leaders the greatest control over rank and file legislators, encouraging party discipline (Mainwaring and Shugart 1997a). As party leaders decide the order of the list, this may also weaken the nexus between legislators and voters. By contrast, in open list systems, as candidates of the same party compete against one another, they face incentives to form factions: that is, organized groups within parties that compete for control of valued resources (Cox and McCubbins 2001). To sum up, while closed list systems encourage party votes, in open list systems legislators face incentives to cultivate personal votes (Ames 1995b).21 The overall degree of fractionalization of the legislature also depends on the incentives for coalition formation and stability generated by the electoral rules. Usually, coalitions will form in the pre-electoral stage in single-member districts, while coalitions will form in the post-electoral stage in multimember districts. However, there is much variation within the multimember district family, as some systems may encourage coalition formation at the pre-electoral stage as well. The binomial system used in Chile is a case in point.22

A word of caution: the link between legislators and voters in open list systems should not be overestimated. In large multimember districts, as individual legislators are encouraged to focus on narrow constituencies, it is more costly for the voters to become informed about the contending candidates, and thus it may be more difficult to hold them accountable (Payne and others 2002). 22 The binomial system is a proportional representation system with district magnitude of two in all districts. Each of the lists receiving the two highest vote shares wins one of the two available seats per districtâ&#x20AC;&#x201D;unless the most voted list outplaces its second place rival by a ratio of more than two to one, in which case both seats go to the most voted list. In addition, coalitions are allowed only to the extent that they are national in scope (thus they are binding in every district in the country). 21

Carlos Scartascini

The binomial system provides parties a strong incentive to coalesce at the district level, as doubling the rivalâ&#x20AC;&#x2122;s vote share gives the winning list all seats, or securing a second place and avoiding being doubled gives the list half the seats. Additionally, given that the costs of defection for a party (such as leaving the coalition) are high, the binomial system also fosters coalition stability, as chapter 5 on Chile in this volume shows. The binomial system seems to be one of the main characteristics that explains some of the good features of public policy in Chile. In contrast, multimember systems, where the thresholds of success are lower, make parties compete unilaterally, thus limiting their ability to form coalitions at the pre-electoral stage. As the related country chapters show, in these cases, the stability of coalitions varies across countries and across administrations; it depends at least in part on the extent to which there is a match between what the president is able to offer legislators to keep them within the coalition, and what legislators want. In the case of Ecuador, agreements are short-lived because legislators do not want to be associated with the president. In the case of Brazil, presidents use the distribution of projects with local benefits to gather political support. Electoral rules also play a role in explaining the representation of ideologies and the polarization of the political system. Thus they also affect legislatorsâ&#x20AC;&#x2122; preferences. The electoral system can affect represenâ&#x20AC;&#x2018; tation of ideologies, minority representation, and representation of local interests (localism), and whether politicians could compete successfully at the local or the national level (Grofman and Reynolds 2001) through their impact on barriers of entry and representation biases.23 While some electoral systems favor the entry of small parties, others require a large presence in the electorate in order to win representation (high thresholds). Similarly, some electoral systems could overrepresent some parties and underrepresent others according to the distribution of their supThe relationship between individual candidates and the party can also be changed through the practice of gerrymandering: dividing a territorial unit into election districts to give one political party an electoral majority in a large number of districts, while concentrating the voting strength of the opposition in as few districts as possible. Gerrymandering can affect the stability of legislators and their political allegiance, and thus it can affect the game of political transactions. 23

Who’s Who

in the

PMP

porters across districts (Saiz and Geser 1999; John and Saiz 1999; Calvo and Murillo 2004) and the degree of malapportionment (Samuels and Snyder 2001; Ansolabehere, Snyder, and Ting 2002).24 Electoral systems affect polarization through electoral formulas and district magnitudes (Sartori 1976; Mainwaring and Scully 1995a). In single-member district (SMD) electoral systems, policies are hardly ideological and a successful politician responds to the preferences of the median voter in his or her geographic district. By contrast, in regimes of multimember districts and proportional representation, parties are more ideologically oriented than in two-party systems and political parties’ preferred policies usually deviate from that of the median voter. A higher degree of polarization could imply the existence of more extreme positions in government, the legislature, or the political spectrum. In some instances, this could lead to greater instability in the political system and more drastic changes in policies (Haggard 2000). Some of the actions of legislators are explained by their expected duration, which affects the rate at which they discount the future (dis‑ count rate). Higher discount rates reduce the value of future benefits and reduce legislators’ incentives to invest in their capabilities while in office. The duration is determined, among other factors, by term limits, reelection constraints, and the electoral system in place. The implications of the first two factors on duration are straightforward; the shorter the term and the harder it is to be reelected, the higher the discount rate. The electoral system affects duration through its impact on determining who has the power to reward or punish the legislators, that is, who is the legislator’s principal.25 In single-member districts such as in the Malapportionment is the inequitable or unsuitable apportioning of representatives to a legislative body. Some districts could be electing more than an equal share of legislators according to the population of the district. Even if malapportionment does not affect the number of relevant players, it can affect the legislator’s identity, and thus the results of the policymaking process. 25 The principal-agent problem is a particular description of a situation under game theory. There is a player called a principal, and one or more other players called agents with utility functions that are in some sense different from the principal’s. The principal can act more effectively through the agents than directly, and must construct incentive schemes to get them to behave at least in part according to the principal’s interests. 24

Carlos Scartascini

United States, it is the voters who are responsible for getting legislators reelected. Since seniority plays an important role in committee assignments, which in turn affect the ability of legislators to deliver benefits to their constituents, as a general rule voters have an incentive to reelect incumbent legislators. In contrast, in a country such as Argentina where the direct connection between voters and legislators is weaker, governors and other provincial party leaders tend to be responsible for putting together the party lists. They tend to be the â&#x20AC;&#x153;principalsâ&#x20AC;? of the legislators, and often have incentives to move legislators to other (often subnational) positions, which results in increased turnover in Congress. Some partial data support this hypothesis. The high rate of reelection of members of the U.S. Congress is not that common in Latin America legislatures. In Argentina, the number of incumbents running for reelection is close to 25 percent, and the reelection rate is below 20 percent (Jones and others 2002). In Chile, however, around three-quarters of incumbents have been renominated for the legislature within the same coalition, and three-fifths have won reelection (Carey 2002a). In the mid-1980s, the percentage of representatives seeking reelection in Argentina was as low as 26 percent, compared to 99 percent in the United States. For the same elections, the percentage of representatives returning to office was 17 percent in Argentina (proportional representation, closed lists) and 83 percent in the United States (plurality). These percentages were 70 and 43 in Brazil (proportional representation, open list), and 76 and 59 in Chile (binomial) (Morgenstern and Nacif 2002; IDB 2005). As shown, the method of election, along with other complementary institutions, is relevant in this framework because it could have important implications for the number, characteristics, preferences, and stability of agents and groups that are in charge of policymaking.26 Additionally, the electoral system could affect the pool of candidates willing to enter politics and the personal characteristics of individual legislators (such

In addition to the electoral system, other determinants that could affect the personal characteristics (competence, honesty) of legislators are candidacy requirements, nomination procedures, term limits, and party organization. Additionally, gender and ethnic constraints could affect entry into politics. 26

Whoâ&#x20AC;&#x2122;s Who

in the

PMP

as their competence and honesty) because different electoral systems affect the incentives faced by individual legislators to extract rents or engage in corrupt activities (Caselli and Morelli 2004).27 According to standard criteria by which legislatures are compared, Chileâ&#x20AC;&#x2122;s congress is reestablishing itself as an unusually professional and technically competent legislature (Carey 2002a; Saiegh 2005; IDB 2005). The same cannot be said of other legislative bodies in the region, where professional capacity is not common and politicians rotate without accumulating any institutional knowledge. Legislative Structure: Unicameralism and Bicameralism

One of the most important characteristics regarding the structure of legislatures is whether they have a single chamber or are divided into two different chambers. In Latin America, more than half the countries have bicameral legislatures (Llanos 2003a; Sanchez, Nolte, and Llanos 2005). Nonetheless, there is little agreement in the literature on the benefits of having a bicameral legislature. While political philosophers like Montesquieu were in favor of the institution, others like Jeremy Bentham were not. Some of the advantages of having a bicameral legislature include the following: avoiding bad decisions made in haste;28 avoiding actions that favor narrow interests (Buchanan and Tullock 1962); reducing the likelihood of voting cycles (Mueller 1996a); and representing different interests (Mueller 1996b).29 Of course, the validity of these arguments

First, while in proportional representation (PR) systems incumbent party leaders monopolize control over rents, in plurality systems the locus of rents is more evenly divided between the party leadership and individual legislators. Second, the ability of voters and opposition parties to control rent extraction under both systems is different. As plurality rule produces districts with smaller numbers of voters than proportional representation systems, collective action problems for voters and opposition groups in monitoring corrupt incumbents are less severe. Thus some authors argue that proportional representation systems are more susceptible to corruption relative to plurality systems (Rose, Ackerman, and Kunicova 2002). 28 This was one of the reasons James Madison, John Jay, and Alexander Hamilton (1787) mentioned in The Federalist Papers. 29 Of course, this begs the question of which interests these may be. 27

Carlos Scartascini

depends on the implementation of the system. In some cases, bicameralism could make the government incapable of responding decisively to a crisis, could increase wasteful redistribution (either to a geographic or to a corporate constituency), or could excessively increase the cost of reaching a decision (Mueller 1996b). In other cases, having a second chamber could provide the necessary balance of power. Because the strategies of the actors and the outcomes are affected by the presence or absence of a second chamber, the study of the policymaking process in countries with bicameral legislatures must focus not only on the bargaining of each one of the chambers separately, but on the interaction between chambers, as well. Two institutions related to the legislature determine each chamberâ&#x20AC;&#x2122;s strength or weakness, its relevance, and the basis of its representation. The first important aspect is the constitutional powers of each chamber (such as the order of voting, rules to overturn/modify decisions, and role in the confirmation/impeachment processes). Those powers determine the role of each chamber in the policymaking process, which one has the most power, and under which conditions both are relevant policy players. Constitutional powers could determine that in certain countries or policy areas both chambers are relevant, which would make it more difficult to pass new legislation. In other countries (or policy areas) one of the chambers could be â&#x20AC;&#x153;subservientâ&#x20AC;? to the other and would not act as an additional veto player, which would make it easier to pass new legislation. The second important feature is the method of election. While this chapter has already discussed the impact of this feature on the role of the legislature as a whole, here the emphasis is on the differential impact on a second chamber. On the one hand, the method of election could affect its political clout. For example, a second chamber that is not directly elected could lack the democratic legitimacy, and thus the real political influence, that popular election confers.30 On the other hand,

Currently in Latin America, there are no cases of indirect election. In Chile, until the constitutional reform of 2005, nine senators were appointed and one seat was reserved for former presidents. 30

Whoâ&#x20AC;&#x2122;s Who

in the

PMP

the electoral rules may affect whether having two chambers adds players and interests to the bargaining, which interests those are, and thus the type of negotiation that ensues. The method of selecting the representatives to each chamber and the basis of representation (type of constituency) determine the degree of congruence between the upper and lower houses in bicameral legislatures. If there is congruence, and the party composition of one chamber mimics the composition of the other, it is often assumed that preferences of the chambers will be similar or identical. However, congruence (and similarity) should not be equated with identity of positions. Legislators in each chamber could be representing different constituencies, as some geographic entities could be overrepresented in one of the chambers. Moreover, chambers could have different decision-making rules. In addition, opinions may vary even within the party. One factor that can explain the divergence of positions across chambers is the degree of the legislatureâ&#x20AC;&#x2122;s malapportionment. Usually, the second chamber is elected by methods designed so as to overrepresent certain minorities. The greatest degree of overrepresentation occurs when there is equality of state representation regardless of statesâ&#x20AC;&#x2122; population, as is usually the case in the upper chamber of federal governments.31 A high degree of overrepresentation (malapportionment) usually produces two effects. First, state-level interests are favored in public policy, particularly on fiscal policy. Second, smaller states usually end up relatively better off. For example, the Brazilian and Argentinean congresses overrepresent the (mostly poor) less populated states (Samuels 2003; Gibson and Calvo 2000). Therefore, even under congruence, the analysis indicates that compared with unicameralism, bicameralism could increase the number of relevant players, and thus make changes to the status quo more difficult (Tsebelis and Money 1997). If changes occur, they happen through a process of both cooperation and conflict between the two chambers. The outcome of the bargaining between the chambers depends on the Numerous explanations for this arrangement have been offered, both normative (in terms of equity) and positive (such as who held the power at the moment of drafting the constitutions). 31

Carlos Scartascini

relative power of each house, which is a function of the constitutional powers and institutional rules (such as in which chamber bills are introduced, which chamber has the last word on disputes, and the number of possible iterations for considering a bill) and the impatience of each chamber to reach a deal. We now turn to those factors. Legislative Organization: Committees and Agenda-Setting Power

In a context where the legislature has an impact on policymaking, analysis of the practices and regulations that rule legislative activities matters because these practices and regulations may alter the number of players and their allegiance. By distributing power and resources, the voting rules, agenda-setting powers, rules for introducing bills, order of voting, presence or absence of roll call votes, and the relevance of committees and seniority all affect the actual number of agents that have influence over policy decisions, their incentives, the arenas in which they interact, and their discount rates. The legislature rules are usually endogenous because politiciansâ&#x20AC;&#x201D;mainly party leadersâ&#x20AC;&#x201D;react to the deeper institutions commonly found in the constitutions, such as the electoral rules, by trying to shape the workings of the political institutions to their advantage (Shepsle and Weingast 1987; Carey 2006; Cox 2006). One important aspect of legislative organization is the rules that guide the process and structure of legislation. Agenda-setting power refers to any special ability given to legislators to determine which bills are considered on the floor and under what procedures.32 Because legislative rights and resources are not evenly distributed among legislators, agenda-setting power affects the structure of the policymaking process and the weight of legislators in policy decisions (Cheibub and Limongi 2002). Understanding agenda-setting power is fundamental to fully grasping the micro workings of legislatures.33 First, it provides clues about the disAgenda power is positive/negative when a legislator or party has the ability to ensure/prevent the consideration of bills on the floor. 33 Cox and McCubbins (2005) explain the workings of the U.S. Congress by analyzing agenda-setting rules. For applications of the agenda power framework in other institutional settings, see Figuereido and Limongi (2000); Amorim Neto, Cox, and Mc32

Who’s Who

in the

PMP

tribution of “power” in the legislatures. Second, it offers insights on the way majority parties or coalitions control the flow of legislation, and thus influence legislative outcomes (Cox and McCubbins 2005).34 For example, in some countries, the president of the chamber controls which bills will be considered in a legislative year. That prerogative increases his or her power, which he or she can use to foster discipline of fellow party legislators and strike deals with the opposition. The second aspect to analyze is the role of committees. Committee power depends on the rules governing the sequence of proposing, amending, and in some cases vetoing proposed bills in the legislative process. In most countries, committees are not only repositories of policy expertise but also gatekeepers and the point of origin of policies in their respective policy domains, which provides them with disproportionate control over the agenda. Therefore, committees are sometimes powerful and they are able to impose many of their policy preferences (Weingast and Marshall 1988; Shepsle and Bonchek 1997). However, the extent of committee strength, the degree of specialization, and the technical capacities of committees vary widely and are shaped by a number of factors. First, committee rules determine the number and size of committees, thus affecting the supply of committee slots. If the number of committees per legislator is large, legislators are required to serve on several committees at the same time. As time and effort are limited resources, and legislators participate in more committees simultaneously, the level of specialization and the degree to which legislators accumulate policy expertise decreases (Jones and others 2002). Second, the process of committee and leadership assignment also affects specialization. While a seniority system in which legislators serve particular sectoral constituencies in order to be reelected fosters specialization, a partisan distribution of committee and leadership assignments, where party leaders practice rotating legislators from one Cubbins (2003); and Jones and Hwang (2005). Tsebelis (2002) also analyzes agenda setting in parliamentary systems. 34 Agenda-setting rules can usually be found in a legislature’s formal procedures (“reglamentos”). Alemán (2006) provides a survey of the internal rules of procedure in Latin American legislatures.

Carlos Scartascini

committee to the other, undermines it. Finally, another factor shaping a committee system’s technical capacities is the possession of resources, such as a competent committee staff (Saiegh 2005). In those cases where committees are important for legislative policymaking, if party leaders retain control of committee nominations,35 they could use those nominations to ensure a higher degree of party discipline (Cox and McCubbins 1994). Thus committees have gained importance in countries such as the United States, where party leaders and party seniority determine access to committee membership (Shepsle and Weingast 1987; Rasch 1999). In the case of Colombia (before a recent reform), the party label was not a relevant indicator of allegiance and factionalization was very high. Even campaign financing would flow directly to the factions, out of the control of party leaders. In that context, the only tool party leaders had for retaining some power over the legislators was through the selection of legislators for the committees (which are few and important). Even though party leaders have usually tried to find ways to increase the importance of committees in order to increase their clout with the rank and file of their parties, in some countries they have chosen the opposite path to overcome problematic situations within their parties. For example, parties have chosen to create new positions and new offices within the parties and to increase the number of positions in the legislature, including increases in the number of committees. In Argentina, leaders use assignments to generate support, both within and among parties, and since reelection rates are low, the number of committee posts has grown to provide leaders more posts to offer to pliant legislaThere are other ways in which party leaders can increase party discipline, such as the allocation of party funds for the electoral campaigns of individual legislators or party nomination procedures. Opposition status also fosters party discipline, despite the fact that the opposition’s party leadership controls fewer resources and is thus more vulnerable. Additionally, when there is an opposition party that is willing to block most of the government initiatives, the discipline in the governing party tends to increase. Members of governing parties in presidential systems tend to feel freer to vote against the executive on the assembly’s floor than their counterparts in parliamentary systems. For details on party discipline in Europe, see Gallagher, Laver, and Mair (1992) and Sánchez de Dios (1999). 35

Who’s Who

in the

PMP

tors. The number of committees grew from 26 in the 1983–85 legislature to 39 in the 1993–95 legislature (Jones 2002; Jones and others 2002; Mustapic 2002). All of this has tended to conspire against the role of committees as repositories of technical expertise and policymaking capabilities (Jones and others 2006; chapter 3 on Argentina, this volume). In contrast, in Colombia, legislators can participate in only one committee for the entire four-year term. These features provide high incumbency rates and a high level of specialization. Party leaders are not the only ones who use committees to their advantage. Legislators usually select to belong to those committees that would provide them with the greatest benefits, such as increasing their chance of reelection. Stratmann and Baur (2002) find empirical evidence of different behaviors across legislators for Germany, where half of the parliamentary seats are awarded from single-member constituencies and the other half through proportional voting. The legislators elected from single-member constituencies, regardless of individual expertise, tend to choose those legislative committees that deal with geographically based affairs, while the legislators elected by party lists tend to prefer those committees that deal with broad-based policies and transfers. Finally, regarding the arenas in which transactions take place, an important characteristic of Latin American legislative institutions is that much negotiation and bargaining occurs behind closed doors (Morgenstern and Nacif 2002). In most instances, presidents prefer to shield disagreements with the legislature, as well as the concessions made to the legislature (or to individual legislators), from the public eye. In other cases, when the president’s proposal enjoys strong popular support, he or she may prefer to override any legislative proposal by relying on public opinion and refusing to offer concessions. The Role of the Legislature in the PMP and Congress’s Capabilities

Recent studies have developed classifications or typologies of Latin American legislatures on the basis of variables and concepts like those emphasized above. By focusing on Argentina, Brazil, Chile, and Mexico,

Carlos Scartascini

Cox and Morgenstern (2001, 2002) classify Latin American legislatures as reactive instead of proactive. This implies that while the legislatures rarely initiate legislation, they are often involved in negotiating over policy issues behind the scenes and vetoing or amending executive initiatives. Accordingly, Latin American legislatures are not necessarily powerless or unimportant, and presidents must anticipate what the assemblies may accept and modify their strategies accordingly. The support for the presidents in the legislature varies greatly in Latin America, making it more difficult to establish a pattern of relationship between the two branches over time. While support for the president usually oscillates around half of the members of the legislature in the United States, in Latin America the amplitude is larger. IDB (2005), drawing on Saiegh (2005) and Stein and Tommasi (2005b), develops an index of the policymaking capabilities of congresses. The index attempts to capture the factors that shape the role of legislatures in the policymaking process by focusing on the capabilities of congress as an organization, as well as on some personal characteristics of legislators. The quantitative and subjective variables that make up the index include: the level of confidence of citizens and business in the performance of congress, the average years of legislator experience, the percentage of legislators with university education, their technical expertise, the average number of committee memberships per legislator, the strength of committees, and the extent to which congress is a desirable place for politicians to build a career. The evidence indicates that those countries with high levels of congressional capabilities tend to score high on their features of public policies. The impact of congress in policymaking is not independent of the role of the judiciary. The evidence seems to indicate that congressional capabilities particularly affect policies in the case of judicial independence; otherwise, there would be no regular enforcement of the acts of congress. The Judiciary Branch

The role of the judiciary is framed by the choice of judicial system: that is, whether a country â&#x20AC;&#x153;choosesâ&#x20AC;? to adopt civil law or common law. Com-

Who’s Who

in the

PMP

mon law is the body of customary law, based upon judicial decisions and embodied in reports of decided cases, that has been administered by the courts of England since the Middle Ages and has evolved into the type of legal system now found also in the United States and in most of the member states of the Commonwealth. Civil law, which has been adopted in much of Latin America, is a set of codes that sets forth general rules that are applied and interpreted taking into consideration the “spirit” of the code in an effort to apply to each case the solution that would have been desired by the legislator (Tullock 1997). This distinction is important, as it frames the relative importance of the judiciary branch vis-à-vis the other branches of government. In terms of the framework of analysis of this study, the judiciary may play several policymaking roles.36 It can be an impartial referee, as an enforcer of political transactions among different political actors; this can increase the durability and stability of policies. The judiciary can also be a policy player, shaping policies according to its preferences and/or society’s, and sometimes providing a voice for marginalized or unorganized social sectors. The judiciary plays these roles through its reactive (veto) and proactive prerogatives. This distinction of the potential roles of the judiciary is important because it provides a more accurate depiction, moving beyond the analysis of its actions only as a veto player and highlighting its importance as enforcer of political transactions. These functions are not mutually exclusive, and some are closely connected. The relevance of the judiciary as an actor in the PMP depends on the degree of judicial independence.37 If the judiciary is independent, legislators and the executive must take into account the preferences of the judiciary when making policy. On the contrary, if the judiciary responds to one of the other branches of government, then its actions would

The framework has been adapted from Sousa (2005). Judicial independence has four interrelated dimensions: substantive independence, or the power to make judicial decisions and exercise official duties subject to no other authority but the law; personal independence, or stability of tenure and freedom from intimidation or threats; collective independence, or judicial participation in the central administration of courts; and internal independence, or independence from judicial superiors and colleagues. 36 37

Carlos Scartascini

merely mimic (and probably strengthen) the actions of that branch of government. Among the characteristics of the judiciary that affect judicial independence, several stand out: the extent of budgetary autonomy; the level of transparency and the extent of the use of meritocratic criteria in the process for nominating and appointing judges; the stability of the tenure of judges; and the reach of judicial review powers. Effective judicial independence also depends upon the behavior of other actors, such as whether the president or political parties regularly interfere with the courts. This, in turn, depends upon these actors’ incentives (Sousa 2005).38 Given judiciary independence, the judiciary, as an impartial referee, can play the role of enforcer of political transactions. This offers an additional layer of durability to politicians’ agreements by bounding them to their commitments. Thus, working mainly in reactive fashion, the judiciary can provide a “durability mechanism” that can increase the probability of reaching intertemporal agreements.39 In this framework, related to the work first discussed by Landes and Posner (1975), the presence of an independent court generates intertemporal enforcement of the political agreements undertaken today, increasing the benefits of implementing policy exchanges.40 In other words, an independent judiciary tends to resolve time inconsistency problems (that is, agreements made today have a higher discounted value because they are less likely to be changed in the future) because judges exhibit a pronounced ten-

A topic related to the independence and duration of the judiciary is the existence and stability of, and respect for, the constitution. Whether or not a country has a written constitution, institutions such as the size of the majorities required for amendment and the type of judicial review for constitutional matters would all contribute to determining the durability of rules. 39 By entering the bargaining with veto power similar to the rest of the agents, it could also facilitate cooperation by enforcing the transactions that facilitate long-term agreements (Crain 2001). 40 Crain and Tollison (1979) show that as judicial independence and/or the tenure of judges increases, there are fewer incentives to use other, stricter rules to prevent time inconsistency problems. (In game theory and economics, time inconsistency is a situation in a dynamic game where a player’s best plan for some future period will not be optimal when that future period arrives). 38

Who’s Who

in the

PMP

dency to resolve legal disputes and ambiguities in terms of the expressed intentions of the legislature that originally enacted the law. Therefore, an independent judiciary, even through vetoing new legislation, could be a facilitator of intertemporal agreements. In this context, a longer duration of judges could contribute to increasing the adaptability and stability of policies (Iaryczower, Spiller, and Tommasi 2002). As a policy player, the judiciary branch can act in a reactive way (as a veto player) or in a proactive manner, molding policies according to its preferences by “ruling from the bench” in common law countries, or in civil law countries by interpreting laws according to the constitution.41 For example, in a context where the judiciary is independent and able to veto new legislation, if the legislature and the executive wish to move policies out of the status quo they would have to approve policies that are closer than the status quo to the preferences of the median judge (Shepsle and Bonchek 1997). Additionally, judges (or constitutional courts, in the case of civil law countries) can introduce their preferences into policymaking by ruling on policies and new legislation if their preferences differ substantially from those of the other agents (Cox and McCubbins 2001; Tsebelis 2002).42 In this role, the judiciary could contribute to making policies more public regarding and to ensuring that policies are more inclusive, if judges’ preferences are aligned with those of the population at large. The preferences and quality of the judges, which affect which interests they represent, are usually determined by the appointment procedure and the rules by which the judiciary is organized.43 To predict their preferences and potential effects on the policymaking process, it is important to understand who judges are and what they want. Following the

The judiciary can become a policy player by interpreting the statutes, not only in terms of what the legislature wrote at a particular time but also in light of the entire legal precedent. 42 That is, if they are not “absorbed,” in terms of the definition by Tsebelis (2002). 43 Tsebelis (2002) discusses appointment procedures and impact on the preferences of the courts. Other factors that shape the judiciary’s independence are the degree of judicial budget autonomy, the terms and tenure of judges, and the extent of judicial review powers (Sousa 2005). 41

Carlos Scartascini

work of Posner (1993), some of the motivations for judges are popularity, prestige, and reputation: popularity among fellow judges; prestige in the legal and larger political community; and reputation in the academic legal world (for example, the desire not to be reversed by a higher court or the legislature). Therefore, in this context, the method of nomination and appointment (by the president, by the legislature, by both of them, by a judicial council, or by some other means) makes the difference. Summarizing, while the judiciary can constitute an additional veto player, making it harder for the government to change policies, or respond to crises, an independent judiciary can also favor the development of political transactions and move policies toward their (or societyâ&#x20AC;&#x2122;s) bliss point.44 On one hand, it can provide enforcement for the agreements reached by other actors. This can increase the durability of agreements and policies, and thus the present value of cooperation: in this case, increasing the adaptability and stability of policies.45 On the other, by acting as a policy player, the judiciary can move policies toward increasing public regardedness, if its incentives are right. Federalism and Subnational Authorities

The basis of federalism is a national polity with dual (or multiple) levels of government; each level exercises exclusive authority over constitutionally determined policy areas, but only one level of governmentâ&#x20AC;&#x201D;the central governmentâ&#x20AC;&#x201D;is internationally sovereign (Gibson 2004). Federalism affects policymaking through the role that subnational authorities (governors) may play in the design and implementation of public policies and their interaction with national-level actors (presidents, legislators). For example, the introduction of subnational elections significantly changed political and party dynamics in Venezuela. The extent to which The bliss point is the point of maximum utility: the point that everyone wants to reach to maximize their utility according to their preferences. 45 If there is no judicial independence and duration is low, actors would have to opt for other means to increase the durability of laws, such as introducing constitutional amendments and qualified majority rules. If this were the case, policies would tend to be more rigid. 44

Who’s Who

in the

PMP

subnational authorities influence national policymaking depends on a number of institutional variables, to which the discussion now turns. The first important institutional variable is the method of selecting subnational authorities. If governors are popularly elected, they play a more important role than if they are appointed by the central government. If governors can be reelected, they have incentives to cater to their constituency and pursue regional goals at the expense of national objectives (Monaldi 2005). Another important institutional variable is the existence of a terri‑ torial chamber in the national congress, as discussed in the section on bicameralism. Most federal countries allow for the representation of territories (states, provinces) in the national policymaking process through senates. These chambers tend to increase the power of subnational political actors, as they provide an additional veto point in the political system. This means that subnational interests need to be taken into account when national actors design and bargain over policies. Another key factor is the overrepresentation of subnational units in the national legislature (malapportionment). Malapportionment strengthens the political power of the least populated states relative to the most populated units. Malapportionment is not a unique feature of territorial chambers. Several lower houses in federal systems show a certain degree of overrepresentation, even in population-based lower chambers.46 As a consequence, overrepresented units may skew policies in their favor, and they typically receive higher resources per capita. This is the case in Argentina, which has the highest level of malapportionment in the upper chamber in Latin America and the third highest in the lower chamber (Samuels and Snyder 2001). The method used for selecting candidates to the national legislature is important because it shapes the incentives of representatives once in office—and more generally, their political careers. If the candidates’ names and order are decided at the subnational level, the potential for regional party leaders to use their influence and resources to influence This is a result of the existence of lower and upper limits to the number of deputies that a certain region may have, among other factors. 46

Carlos Scartascini

the election of legislators is high. In that case, the regional congressional delegation to the national legislature might vote more according to the governor’s line than to the national party leader’s or president’s line, giving another tool to the regions for influencing national politics. Moreover, when subnational authorities (such as governors) are important political players at the local level and parties are organized along territorial lines, national elections are heavily influenced by subnationallevel politics, as governors may provide legislators and presidents with electoral coattails (Jones 1997; Samuels 2003). In cases where electoral districts coincide with territorial units (such as states and provinces), the degree of “partisan harmony” (the extent of support for the president throughout the territorial units) affects policymaking. For example, when national leaders lack support at the subnational level, a “vertically” divided government, combined with the absence of sufficient resources to buy support, could complicate the approval of a president’s agenda (Rodden and Wibbels 2002). In Argentina, it contributes to sustaining a political system that operates on the basis of exchanges of provincial support of national policies for fiscal benefits to the provinces. Finally, the link between federal and subnational politics is also fiscal. In particular, federal fiscal arrangements define tax and expenditure assignments between different levels of government, the design of intergovernmental transfers, and the borrowing autonomy of subnational units (Stein 1999). In situations of high fiscal decentralization, subnational authorities control resources that render them powerful actors, even affecting policy outcomes at the national level.47 In Brazil, in the early 1990s, governors would challenge the central government fiscal sustainability to gain leverage in their negotiations (IDB 2005). For example, in many federal countries, while expenditure is decentralized, most revenues are collected at the center and then transferred to the subnational governments. This creates an incentive for subnational governments to overspend from the common pool of resources, enjoying the full benefits of overspending without internalizing its costs. If this tendency is not limited by the central government, the opportunistic behavior of subnational governments may result in dire economic consequences: excessive spending, fiscal deficits, debt crisis, and difficulties in macroeconomic management and fiscal adjustment at the national level (Wibbels 2000; Rodden and Wibbels 2002). 47

Whoâ&#x20AC;&#x2122;s Who

in the

PMP

Political Parties and Party Systems

Political parties are organizations that seek influence in a state, often by attempting to occupy positions in government by â&#x20AC;&#x153;aggregatingâ&#x20AC;? interests in the society (Ware 1996). They have also been regarded as instruments used by politicians to gain political office by reducing transaction costs (Aldrich 1995). In terms of transactions with citizens, political parties reduce information costs by association with party labels and ideologies. In terms of transactions within the political system, political parties reduce the number of players in charge of transactions. The role of political parties and party systems in the policymaking process is twofold. In some countries, political parties act directly in the policymaking process by contributing to the definition and articulation of policy programs and engaging effectively in public policy debates. In others, party system characteristics affect the policymaking process indirectly by influencing the workability of executive-legislative relations, the possibilities for coordination in congress, and the incentives of elected officials: that is, the extent to which they focus on adopting and implementing public policies consistent with a broader public good. The characteristics identified in the literature, especially those that are most relevant for the focus of this volume, include the degree of party system institutionalization, the programmatic character of parties and party systems, the degree of fragmentation, the level of party discipline, and the degree of party system nationalization, as well as the nature of campaign finance.48 Party System Institutionalization

Party systems can be considered institutionalized when the patterns of interparty competition are relatively stable, parties have fairly stable and deep bases of societal support, parties and elections are viewed as legitimate and as the sole instruments for determining who governs,

This chapter does not consider the impact of campaign finance because it is not relevant for the Latin American cases included in this book. 48

Carlos Scartascini

and party organizations are characterized by reasonably stable rules and structures (Mainwaring and Scully 1995a). In terms of the framework of this book, institutionalized party systems are likely to promote longer time horizons, greater policy consistency over time, and a greater potential for intertemporal agreements, since commitments made by current party leaders are more likely to be respected in the future. In contrast, electoral competition in noninstitutionalized party systems is volatile, and linkages between parties and voters are weaker, which may result in noncooperative outcomes in the policymaking process and a high degree of policy volatility. Programmatic Character of Party Systems

Programmatic parties compete for and obtain support on the basis of their policy orientations and accomplishments, and parties distinguish themselves in terms of their policy proposals or ideological orientation. Programmatic parties are usually contrasted with clientelistic parties. The latter compete for and obtain support based on the distribution of selective material incentives to voters (such as public sector jobs, governmental contracts, cash, or meals) in networks of direct exchange and are judged by voters primarily on their ability to deliver these particularistic benefits (Kitschelt 2000). These characteristics affect the outcomes of the policymaking process in contrasting ways. On the one hand, if an institutionalized party system is also programmatic, then political parties are likely to favor more public regarding policies, since parties represent different policy options and voters can hold them accountable on that basis. On the other hand, clientelistic parties are mostly interested in maintaining their narrow bases of support and keeping their electoral machines running. Thus the usual outcome of exchanges between clientelistic parties are private regarding policies. Party System Fragmentation

The degree of fragmentation is usually captured through the number of parties that regularly obtain a significant share of the votes and/or seats

Whoâ&#x20AC;&#x2122;s Who

in the

PMP

in the legislature (Lijphart 1994). In terms of the framework developed in chapter 1, the level of fragmentation could influence the number of players in charge of policymaking. In presidential contexts, it limits the size of presidential legislative contingents and increases the number of partners with which the president must form coalitions. Higher fragmentation would be expected to complicate executive-legislative relations, increase the transaction costs of obtaining policy agreements, and limit policy adaptability. As mentioned, party system fragmentation is a function of the electoral system: the proportional or majoritarian design of the system, the size of electoral districts, the nature of the formula for converting votes into seats, and the concurrence of presidential and legislative elections. However, electoral rules do not affect fragmentation in isolation. In fact, the number of parties is determined by the interaction of electoral systems, the number of salient social and economic cleavages, and the political history of each country (Cox and Amorim Neto 1997). Party Discipline

Party discipline corresponds to the extent that representatives of the same party vote in similar ways in the assembly (Rasch 1999). As mentioned, party discipline is a key factor shaping the presidentâ&#x20AC;&#x2122;s ability to pass his agenda, and it depends on several institutional configurations. For example, it is partly a result of the role party leaders play in nominating and influencing the reelection chances and future political careers of members of congress. However, factors other than candidate selection matter as well, including party leadersâ&#x20AC;&#x2122; roles in organizing the work of the legislature (such as appointing committee members and chairpersons, and agenda power). Party System Nationalization

Another dimension of political party systems is their level of nationalization: that is, the extent to which parties are national in scope and receive similar levels of support throughout the country (Jones and Mainwaring

Carlos Scartascini

2003). Nationalization matters for policymaking because it affects the number of players interacting in the PMP, executive-legislative relations, and thus, the outer features of public policies. When a party system is said to be nationalized, the executive (generally a nationally oriented political player) may be able to pass his or her agenda through the legislature more easily by negotiating with a few key national party leaders. The level of nationalization may also affect the quality of public policies. When the territorial distribution of a partyâ&#x20AC;&#x2122;s vote is relatively homogeneous, politicians will be more likely to treat its constituent units in a similar fashion in areas such as fiscal and social policy. Additionally, as national issues (such as macroeconomic stability) are central to the careers of both the executive and legislators, politicians have incentives to work for delivering national public goods, instead of focusing on delivering particularistic benefits. In contrast, in highly denationalized party systems, parties tend to favor their bases of support and may use different mechanisms (the budget process, discretionary handouts) to distribute resources asymmetrically. Bureaucracies

Several characteristics of bureaucracies and public employment are important for policymaking because these characteristics can affect both the quality of implementation and the enforcement of political agreements (Zuvanic and Iacovello 2005). Two characteristics in particular help explain different bureaucratic types: the degree of autonomy (the extent to which effective guarantees of professionalism in the civil service are in place and the degree to which civil servants are protected from arbitrariness and politization); and the technical capacities of bureaucracies (the degree to which the bureaucracy has salary compensation and evaluation systems). Meritocratic bureaucracies are characterized by high levels of autonomy and capacity. In contrast, clientelistic bureaucracies lack both attributes, and thus function mainly as a private source of employment managed by governing political parties. Between these two extremes, administrative bureaucracies enjoy autonomy but lack a high

Whoâ&#x20AC;&#x2122;s Who

in the

PMP

degree of capacity, while parallel bureaucracies are characterized by high degrees of capacity and low autonomy. The advantages of meritocratic bureaucracies for policymaking are numerous. First, an organized civil service can help politicians fortify their commitment by delegating decision-making authority to autonomous institutions,49 reducing the capacity to reverse their decisions in response to short-term considerations. Because the effectiveness of policies depends on the widespread belief that they will be sustained over time, meritocratic bureaucracies, characterized by independence and long tenure of public employees, are an important part of the set of political institutions conducive to policy outcomes. The evidence seems to indicate that better bureaucratic types are strongly associated with better policy features (IDB 2005). However, all bureaucracies are not created equal. In fact, the extent to which bureaucracies enjoy these characteristics depends on several factors. One important determinant of the organization of the civil service is the strategic interaction between other players in the policymaking process (such as legislatures and executives), their time horizons, the degree of interest alignment, and the distribution of the benefits of patronage (Geddes 1991; Spiller and Urbiztondo 1994). Other Institutions

This chapter has focused on the workings of traditional political institutions. However, other relevant institutional dimensions play a role in defining the incentives and behavior of political actors in the design and implementation of policiesâ&#x20AC;&#x201D;particularly in Latin America. This section considers a few examples. A countryâ&#x20AC;&#x2122;s history of the voting franchise along dimensions such as age, gender, literacy, and geography has a potential impact on the economic and political landscape. This legacy may be important in understanding the interaction of voters with political parties, and the workings The analysis is somewhat similar to the rationale behind the literature on central bank independence and independent regulatory agencies. 49

Carlos Scartascini

of party systems, for example. In countries that lack a long-established democratic tradition, the link between voters and parties is weak, and party competition may be volatile. In such contexts, personalism (the personal draw of individual candidates because of their charisma, their background, or their status as a celebrity, for example) plays a much greater role in voting, increasing outsider candidatesâ&#x20AC;&#x2122; chances of reaching high executive posts. An additional element is the role played by institutional interruptionsâ&#x20AC;&#x201D;particularly by the extent and frequency of military governments. Recurrent institutional interruptions tend to reduce the stability of rules and conspire against the development of policymaking capabilities of key democratic political institutions such as legislatures or supreme courts. Regardless of the de jure institution of the country, frequent institutional changes reduce the time horizons of political players, who may only prioritize short-term political benefits when bargaining over policy. Thus it would be more difficult to sustain intertemporal cooperation in this context. Conclusion

This chapter has presented a survey of the political science and political economy literature to highlight the institutions and actors that are important within the context of the framework developed in the previous chapter. As chapter 1 points out, several characteristics of the policymaking game determine the features of policies. These include the number of actors, their incentives, their discount factors, and the arenas where transactions take place. Every one of the institutions presented in this chapter has some impact on these characteristics. In the case of the executive branch, the type of political regime, the extent of presidential powers, and the method of election and selection can affect the number of agents with influence on policymaking, their stability (and discount rate), the availability of enforcement mechanisms, and the arenas where transactions take place. As the power of the president increases, the governmentâ&#x20AC;&#x2122;s capacity to generate changes and new policies tends to increase as well. While a weak executive must work its

Whoâ&#x20AC;&#x2122;s Who

in the

PMP

policies through the legislature, in strong presidential systems the president can try to force policies unilaterally through executive decrees, and political transactions tend to occur in less formal environments. The role of the legislature in the policymaking process is affected by the powers of the president but also by its own institutional framework; particularly, the rules of access to the legislature and the rules that organize the workings of the legislature. These institutions determine the number of relevant political actors and their relative duration by having some bearing on the number of political parties that can compete successfully for legislative seats, the term of the appointments, and the legislatorsâ&#x20AC;&#x2122; discipline to the party, their source of support and allegiance, and their incentives for specializing and building up their capabilities. These characteristics of the political system help to explain the type of negotiations that take place in Congress and whether Congress can become a relevant arena for policymaking. The impact of congress in the policymaking process is not independent of the role of the judiciary. The judiciary can have a role as an impartial referee, enforcing the acts of the other branches, and can have its own role in the policymaking process as a policy player. While the judiciary has usually been regarded as an additional veto player, this book also stresses its role as facilitator of intertemporal cooperation. By providing enforcement to the agreements achieved by other actors, it increases the durability of agreements and policies, and consequently, the present value of cooperation. Thus it could facilitate the adoption of stable and adaptable policies. The comparative analysis of the country chapters shows that independent judiciaries tend to favor those features of public policies. Even though this book focuses on policies at the national level, the organization of government at the local level matters as well, because subnational-level politicians can also influence national policies. The main channel of influence is the capacity of subnational-level politicians (local party bosses) to exercise control over national-level politicians (governorsâ&#x20AC;&#x2122; influence on national legislators, for example). The method of selecting those national-level politicians is usually the main determinant: the degree of influence will be higher if national politicians are elected

Carlos Scartascini

at the local level where local party bosses can influence the selection of candidates. The chapters on Argentina and Brazil show this vividly. Across all the institutions surveyed, the role of political parties is essential to explain the particular dynamics in each branch of governmentâ&#x20AC;&#x201D;and across branches. The degree of institutionalization and nationalization of political parties, their programmatic character, the fragmentation of the party system, and their internal discipline are all characteristics that affect the number of players, their incentives, the arenas where decisions are taken, and so forth. As the country chapters show, the particular characteristics of each party system interact differently in each institutional context, creating a very distinct policymaking process. The framework of chapter 1 stresses the need for a systemic approach, one that emphasizes configurations of institutions and interactive effects. The country cases in this volume will look in great detail into these configurations and interactions, offering a general equilibrium perspective on the workings of political institutions, policymaking processes, and policy outcomes in Latin America. Yet in order to understand these interactions among multiple institutional dimensions, it is important to first understand each of them individually, focusing on the variety of rules in place in Latin America, and the way they affect the incentives of political actors and the way they play the game. For this reason, this chapter has focused on a number of distinct institutional dimensions of democratic systems and studied them one at a time. We hope that the chapter will provide the reader with the tools necessary to embark on the fascinating reading of the country cases that follow.

3 Political Institutions, Policymaking Processes, and Policy Outcomes in Argentina Pablo T. Spiller and Mariano Tommasi *

Introduction

In the 1990s, Argentina undertook a wide and profound process of market-oriented reforms. With its ambitious program of macroeconomic stabilization, financial liberalization, privatization, and deregulation, Argentina became the poster child of the Washington establishment. The cornerstone of that stabilization-cum-structural reform effort was a monetary regime known as â&#x20AC;&#x153;convertibility,â&#x20AC;? which rigidly tied the peso to the dollar at a one-to-one rate. After decades of inward-looking policies, stagnation, and fiscal crises that led to hyperinflation in 1989, Argentina seemed to have found its way. The macroeconomic performance of Argentina for much of the 1990s was very strong. After GDP declined in the 1980s, growth performance was impressive from 1991 to early 1998. Inflation fell from hyperinflation levels in 1990 to around zero in 1997. Despite the promising results in the 1990s, the Argentine economy entered a long recession in 1998, which exploded into one of the deepest crises in modern economic history in December 2001. In the end, the 1990s were just one more episode in the long history of hope and despair This chapter is based on Spiller and Tommasi (2003, 2005, and 2007). We thank two anonymous referees and Ernesto Stein for valuable comments and suggestions. Mariano Tommasi acknowledges the financial support of the John Simon Guggenheim Memorial Foundation. *

Pablo T. Spiller

and

Mariano Tommasi

that characterized Argentina for most of the twentieth century. Most economists who have evaluated the dismal performance of Argentina have pointed to poor economic policies as the culprit of these sad outcomes. We are inclined to agree with that characterization, but instead of blaming the content of economic policies, we blame the characteristics of policies and policymaking, including policy instability, inadequate enforcement, inadequate commitment capacities, and an inability to effect necessary adjustments. Argentine policies are unstable in ways that weaken their credibility in the eyes of economic actors, rendering them far less effective in bringing about desired economic behavior, such as investment, savings, and job creation, and hence desired economic outcomes, such as sustainable growth and employment. Argentine policies are not only unstable but also poorly coordinated within the tiers of the countryâ&#x20AC;&#x2122;s federal structure and among ministries, secretaries, and programs of the national government. In many instances, it is also patently clear that the investments in capabilities required to produce effective policies are absent. The Argentine state is, in essence, ineffective in enforcing its policies. We argue in this chapter that these undesirable properties of Argentine public policies are the result of a noncooperative policymaking process. Historical legacies and the constitutional context make the Argentine congress a weak policymaking arena that lacks professionalization. Legislators respond to provincial party elites who care little about the quality of national policies. Furthermore, presidential proactive powers are too extensive in practice. As a result, the relevant policymaking actors, such as the president, the provincial governors, and interest groups, lack an institutionalized arena in which they can make intertemporal policy agreements. Additionally, they cannot delegate the implementation of potential policy agreements to a professional bureaucracy because the Argentine bureaucracy has several intrinsic weaknesses. Nor can they rely on enforcement of contracts by the judiciary, because it is weak and easily politicized. Therefore, policymaking becomes the outcome of a noncooperative game in which each actor behaves opportunistically and tries to maximize short-term benefits. This causes policy volatility, poor coordination, and poor enforcement. Political and economic ac-

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

tors (both domestic and foreign) distrust the Argentine polity’s ability to deliver credible policies. In order to overcome this credibility problem, policymakers occasionally resort to very rigid mechanisms, such as the convertibility regime. In the face of large adverse economic shocks, these solutions often turn out to be very costly. Public Policies in Argentina

Perhaps the most noticeable characteristic of public policies in Argentina is their instability. One aggregate indicator of policy stability could be constructed from international indexes, such as the Fraser Index of Economic Freedom, which grades a country’s economic policy according to market friendliness. Figure 3.1 plots the value of that index from 1970 to 2003 for a small number of countries. In the mid-1970s, Argentina went from being one of the most market friendly countries in this sample to being the least friendly after the Soviet Union, and then returned to market friendliness during the reform process of the 1990s. Argentina is one of the two countries in Latin America (with Venezuela) where there has been some recent “backsliding” from market-oriented reforms (Lora, Panizza, and Quispe-Agnoli 2004). Looking at the whole sample of 106 countries, Argentina appears as the seventh most volatile according to the coefficient of variation of the Fraser Index over time. Treating countries whose market friendliness goes up and down, like Argentina’s, separately from countries like Chile or Russia whose policies moved in just one direction (toward market liberalization), Argentina ranks as the fourth most unstable. Argentina exhibits policy instability not only at this aggregate level but also in more specific policies. In Spiller and Tommasi (2007, ch. 7), we document the volatility of antipoverty programs in the 1990s, describing how large policy changes were made without congressional mandates. Existing welfare programs are often reshuffled (refocusing or discontinuing existing programs, and creating new ones) when new ministers or secretaries take office, a frequent event in Argentina. Often, this reshuffling involves substantial tinkering with the geographic distribution of funds. The average tenure of department heads at the National

Pablo T. Spiller

and

Mariano Tommasi

Figure 3.1. Volatility of Economic Policies, Selected Countries, 1970â&#x20AC;&#x201C;2003 9 8

Economic Freedom Index

7 6 5 4 3 2 1 0 1970

1975

1980

1985

1990

1995

2000

2003

Argentina

Uruguay

Colombia

Japan

Chile

India

Russia

United States

Source: Fraser Index of Economic Freedom

Secretariat for Social Development is less than a year. Furthermore, the agency has changed from secretariat to undersecretariat to ministry. The position of agency head has been occupied by highly qualified technocrats, by high-profile politicians, by the spouse of one president, and by the sister of another one. Argentina also shows high volatility in international comparisons of fiscal policy (Mody and Schindler 2004). This policy volatility shortens the time horizons of the economic and social actors, thus reducing policy effectiveness and inducing poor economic and social outcomes. Businesspeople indicate in international surveys that they view Argentinaâ&#x20AC;&#x2122;s policy volatility as very costly for the operation of their businesses. According to the Global Competitiveness Report of the World Economic Forum, when businesspeople were asked

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

how costly for the operation of their business were legal changes in the country, Argentina ranked 61 out of 75 countries in terms of the stability of its business environment.1 Argentina’s business community is reluctant to respond to export-promotion policies because their durability is so uncertain (Acuña 1991). Similarly, trade policy uncertainty has had a negative impact on the macroeconomic performance of Argentina. Uncertainty over trade policy was one of the key factors in the poor growth performance of Argentina in the second half of the twentieth century, according to Hopenhayn and Neumeyer (2003). Argentina is also a weak enforcer of its policies, at the bottom of the international list in its ability to enforce tax collection, social security contributions, and payment of minimum wages (Argentina ranks 75th and 70th out of 75 countries in its enforcement of minimum wages and tax collection, respectively). This inadequate enforcement also weakens the credibility of Argentine policies. Unlike the case of other countries, such as Chile, as argued by Cristóbal Aninat and his colleagues in chapter 5, legislation is not a very hard currency for intertemporal policy exchanges in Argentina. The lack of credibility leads, sometimes, to the adoption of highly rigid policies. A notable example was the 1991 adoption of the straightjacket monetary mechanism of convertibility, which prevented the authorities from undertaking monetary policies in response to adverse shocks throughout the 1990s. It is paradoxical that one of the few crosscountry indicators in which Argentina looked good in the eyes of the international business community was exchange rate stability. An April 2001 survey (just eight months before the convertibility regime fell and the currency suffered a major devaluation) asked businesspeople whether they expected the exchange rate to be volatile, and Argentina ranked the sixth best country by this measure. After the fact, any measure of exchange rate volatility would have placed Argentina as the most volatile country in the sample. This suggests that credibility can be temporarily

For details and references on all the statistics referred to in this section, see Spiller and Tommasi (2007, ch. 7). 1

Pablo T. Spiller

and

Mariano Tommasi

achieved only through very rigid mechanisms. These mechanisms, however, may become very costly under some circumstances. Another example of Argentina’s inability to undertake efficient and flexible policies is provided by the history of fiscal federalism in the country (Tommasi 2006). In attempts to protect themselves from the opportunistic behavior of other political actors, national and provincial authorities have introduced all kinds of rigidities into the federal taxsharing agreement. These rigidities, such as tying specific tax revenues to specific geographic distributions, created numerous microeconomic inefficiencies during the 1980s and 1990s. One of the most recent examples was the 2000 agreement by the national government to transfer fixed nominal amounts to the provinces. In the great economic downturn of 2001, with the convertibility straightjacket in effect and the reluctance of creditors to absorb fresh debt, it became impossible for the national government to honor those intergovernmental commitments. The political disputes over distribution of available fiscal funds and the state’s inability to adjust the fiscal arrangement to the new adverse circumstances are seen as the proximate causes of the Argentine default and crisis of 2001–02.2 To give some texture to the description of policy characteristics, we summarize here one of the various examples analyzed in Spiller and Tommasi (2007, ch. 7): international trade negotiations. International trade negotiations are a crucial policy area for Argentina because access to international markets may hold the key to the country’s development. Our summary of policymaking in this area draws on a recent assessment by one of the foremost Argentine experts in the field, Roberto Bouzas (Bouzas and Pagnotta 2003). As highlighted by Bouzas and Pagnotta, in international trade negotiations and international trade policy more generally, intertemporal capabilities—consensus building, policy consistency, capacity building, and institutionalization—are particularly important. These features, however, are lacking in Argentina. Argentina participates in multiple trade negotiations with MERCOSUR, the World Trade Organization, and the European Union, and 2

See Eaton (2005) for references to international financial press accounts.

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

within the Americas in the framework of the Free Trade Area of the Americas. Bouzas and Pagnotta (2003, p. 81) describe Argentina’s strategy this way: The strategy of external commercial negotiations by Argentina has had diffuse aims and has been ambiguous. The negotiating position has been guided by multiple, and often contradictory, priorities and interests. This is the result of the absence of a domestic consensus that ensures the continuity of those aims over time. Worse still, even for those topics on which there is consensus about the importance of the negotiation (as in the case of participation in the OMC), the ability to act in a consistent way has been very limited.

In terms of institutional organization, Bouzas and Pagnotta (2003) describe the design and execution of international commercial negotiations as highly fragmented, with diverse public offices having overlapping functions and with little coordination. The Ministry of the Economy is the executive office with the most extensive responsibilities, which include the design and in some cases application of several instruments of trade and related policies. Several secretaries under the Ministry of Production are in charge of the application of some of the instruments, and also participate in international commercial negotiations. The Ministry of Foreign Affairs is the main office in charge of international commercial negotiations and it also designs and implements commercial promotion policies. Other agencies linked to the executive also active in the area of trade policy are the National Commission of Foreign Trade, the Federal Administration of Public Revenues, the Investment and Foreign Trade Bank, and the Export-Ar Foundation. This complex organizational structure requires a high degree of coordination. Instead, according to Bouzas and Pagnotta, “there is no formal deliberative process in the elaboration of these policies.” They go on (2003, pp. 91–92): The participation of Congress is sporadic and reactive. Although coordination inside the executive branch has changed over time

Pablo T. Spiller

and

Mariano Tommasi

… the overlapping of functions, competition across offices, and interbureaucratic struggle have been the rule. This means that the efficiency of the routines and procedures depends strongly on particular circumstances, leadership, and personalities … The bureaucratic structures are unstable, and there is high turnover among civil servants, as well as a low level of institutional learning. The high proportion of personnel under temporary contracts conspires against the accumulation of expertise, precisely in an area in which knowledge constitutes a strategic asset … The institutional fragility of the policymaking process spreads to the absence of mechanisms for the systematic and organic participation of the private sector, which has promoted lobbying, in a context of a lack of transparent procedures and a general absence of routines.

Bouzas and Pagnotta’s description of policy features and policymaking in international trade makes numerous points that are useful for the broader concerns of this chapter, and that are consistent with the insights glimpsed from several other policy areas in chapter 7 of Spiller and Tommasi (2007), including social policies, fiscal federalism, and regulation of utilities. We provide below a selected summary of the characteristics highlighted by those policy examples, which we take as “stylized facts” to be explained. The characteristics of the policymaking process illustrated include: deficiencies of the budget process, which gives excessive discretion to some executive actors; insufficient involvement by congress; instability of the bureaucratic structures in charge of implementation; instability of top bureaucratic personnel; noncooperative interactions between national authorities and provincial authorities; noncooperative interactions among and within national ministries; provincial governors who appear as relevant actors in national policy; and promises that are not fulfilled. Policies in Argentina, then, seem to present the following characteristics: They show a high level of volatility (they often change with low-level political rotation). Paradoxically, other policies are too rigid, not being adjusted when circumstances change. Sometimes they respond to changing circumstances in the wrong direction. They are poorly coordinated and

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

poorly enforced, and they present poor qualities in dimensions beyond those captured by the previous points.3 All these policy characteristics are consistent with the predictions of the theory summarized in chapter 1 for cases in which the equilibrium of the game of policymaking is characterized as noncooperative. In the rest of this chapter we describe the workings of the political system in Argentina and we show that those interactions end up generating the conditions that the framework identifies as conducive to noncooperative policymaking. The Policymaking Process in Argentina

In a nutshell, the national policymaking arena of Argentina is populated by actors who have little incentive and few instruments to engage in intertemporal policy agreements. In such an environment, noncooperative short-term strategies are the norm, leading to the policy characteristics described above. The framework presented in chapter 1 suggests that efficient and effective intertemporal transactions require self-enforcement and/or institutional arrangements that facilitate enforcement. In the absence of either type of enforcement, political and policy actions will be characterized by short-termism, inflexible rules, inability to implement efficient policy changes, and underinvestment in capacities, all leading to lowquality policies. Argentinaâ&#x20AC;&#x2122;s historical political instability, basic constitutional features, electoral rules, and federal fiscal features are key determinants of its inability to develop efficient long-term public policies.

In this section we have painted a fairly bleak picture of Argentine policies. Several indicators from international comparisons are consistent with our statements. There are other indicators in which Argentina does not fare so badly in global international comparisons. Yet the message of this paper is that Argentina has policies of very low quality, especially when compared with obvious determinants of policy quality, such as human capital or level of development. Berkman and others (2007) build an international index of policy quality. Argentina appears in the bottom 40 percent of countries in that index, but it moves down to the bottom 20 percent of countries when that index is adjusted by GDP per capita. 3

Pablo T. Spiller

and

Mariano Tommasi

The theoretical discussion of chapter 1 identified various elements that affect the capacity to achieve efficient intertemporal exchanges. These elements include the number of political actors with power over a given decision, their intertemporal linkages (how long they are in office), the characteristics of the arenas in which they undertake their political exchanges, and the availability of enforcement technologies—such as an independent and capable bureaucracy or an independent supreme court. Most of these features of policymaking “take the wrong values” in Argentina. Argentina’s key political actors have tended to have short political horizons. The unusual political instability of the twentieth century contributed to that history. From the 1930 coup until 2007, there was only one transfer of power from one president to the next (from Menem to De la Rúa in 1999) that strictly followed constitutional norms. Political instability has left an imprint through path-dependent behavior in congress, the courts, the bureaucracy, and the federal fiscal system, as well as through the actions and expectations of nongovernmental actors. Political instability, however, is not the only factor contributing to shortsighted behavior. Electoral rules that transfer power away from congress and national parties toward provincial political patrons (who are not particularly interested in building a strong national congress) contribute to the shortening of legislators’ political horizons, and, in an institutional general equilibrium way, affect the incentives of the rest of the polity. The ability to enter into efficient intertemporal political agreements has also been eroded by weak constraints on unilateral moves by some actors. For instance, weak constitutional, judicial, and budgetary constraints on moves by the executive have led to unilateralism on the part of the president, which in turns weakens the incentives to achieve congressional agreements in the first place. A similar logic has been at work in intergovernmental relations: weak constraints on some moves by the national government on issues that affect the provinces have led to noncooperative behavior in the federal fiscal game (see the discussion on fiscal federalism below). The history of political instability has contributed to the lack of judicial, and thus constitutional, constraints on executive action. A profes-

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

sional bureaucracy, well supervised by congress, could be an alternative channel for the intertemporal enforcement of political agreements. But Argentina does not have such a bureaucracy, either: in part because of its history of instability, but also because of current incentives of key political players. The combination of lack of legislative incentives, the ability of the executive to act unilaterally, and the power of provincial leaders all have moved crucial political and policy bargaining away from the national legislature and into other arenas. Some key policy decisions take place in executive quarters (among the president, a key minister, advisers, and a few businesspeople), in meetings of the president with some governors, or in other closed groups. Not only are those arenas not transparent, but they also lack the required institutional stickiness to enforce bargains over time. The Argentine policymaking process can, then, be summarized in a number of propositions about the behavior of some of the main institutional actors and about the characteristics of some of the main institutional arenas. We set forth these propositions below, and we devote the rest of the chapter to explaining and supporting them. 1. Congress is not an important policymaking arena. 2. The executive tends to have substantial leeway to take unilateral policy action. 3. Provincial political powers (especially provincial governors) are very important in national policymaking. 4. There is a symbiotic interaction between national and provincial policymaking that operates through political and federal fiscal channels. 5. Considerations of fiscal federalism are a factor in almost every policy issue, adding transaction difficulties and rigidities to policymaking. 6. Given the incentives of the executive of the day, of legislators, and of provincial governors, there is little investment in policymaking capacities in several spheres.

Pablo T. Spiller

and

Mariano Tommasi

7. By and large, the bureaucracy is not an effective corps to which to delegate the technical implementation of policy bargains. 8. The judiciary does not provide much intertemporal â&#x20AC;&#x153;glueâ&#x20AC;? to political or policy agreements. 9. Nongovernmental actors in the policy process (such as business groups and unions), lacking a well-institutionalized environment for political exchange, usually follow strategies that attempt to maximize short-term benefits.

The configuration of and interactions among the factors mentioned above is what, in our view, explain the characteristics of Argentine policies described in the previous section, and we explain those interactions in the rest of the chapter. We can, nonetheless, anticipate some direct policy implications arising from these propositions. The fact that the executive has wide leeway to take unilateral policy decisions means that policy is unstable and lacking in credibility. The importance of provincial political powers makes fiscal bargains particularly difficult to strike, with the consequent implications for the quality of fiscal and macroeconomic policy. One of the features of the Argentine bureaucracy is a high rate of turnover in top positions, including many outsiders who pass briefly through the public sector (having little more attachment than loyalty to the political patron who places them there). This turnover limits the extent of institutional knowledge and the development of cooperation across ministries and secretariats, deepens the heterogeneity in policy quality, and reinforces the lack of policy coherence. In the rest of the chapter we argue that the deep determinants of the derived institutional characteristics and behavior summarized in the propositions above are some basic institutional characteristics of Argentina (including prominently some aspects of electoral rules), as well as some historical legacies at the return to democracy in 1983. We also argue that several of these characteristics tend to reinforce one another over time. For instance, the lack of judicial enforcement of previous political agreements reflected in laws or in the constitution diminishes the value of legislation and hence the political value of being in congress. This, in turn, affects legislatorsâ&#x20AC;&#x2122; incentives to invest in strengthening congress or

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

in building long-term legislative careers. These congressional weaknesses reinforce the need and the temptation of the executive to undertake unilateral actions to deal with policy issues, and so forth. In the next section, we provide a brief introduction to Argentinaâ&#x20AC;&#x2122;s political institutions and history, and in the rest of the chapter we bring some of the components of that picture into focus. A Brief Introduction to Political Institutions in Argentina

Like the United States, Argentina is a federal republic, with a presidential form of government and a bicameral legislature. Given these similarities in some basic constitutional features, and the fact that the U.S. political system is the one most studied by positive political theorists, we will follow the expositional device of describing some features of Argentine institutions and Argentine political behavior by contrasting them with their U.S. counterparts. A key difference between the political organization of the United States and Argentina is the way legislators are elected. The members of the Argentine chamber of deputies (currently 257) are elected from multimember districts (the 23 provinces and the capital city) for four-year terms. The deputies are elected from closed party lists. Half the chamber is renewed every two years, as each district replaces half its legislators (or the closest equivalent). Unlike in the United States, where each state is represented by a number of U.S. representatives proportional to its population, in Argentina the less populous provinces are highly overrepresented in the chamber. While chamber seats are technically allocated among the provinces on the basis of their population, each province has a minimum of five deputies. For ten provinces, this truncation leads to overrepresentation. Overrepresentation is even higher in the upper chamber. Before the 1994 constitutional reform (and as in the United States), each district was represented by two senators. Unlike in the United States, however, senators were elected indirectly for nine-year terms by the provincial legislatures using the plurality formula. Since the constitutional reform of 1994, the senate is composed of 72 members, with every province

Pablo T. Spiller

and

Mariano Tommasi

represented by three senators (elected directly since 2001 using the incomplete list electoral rule), with the stipulation that no one party can occupy more than two of a province’s seats in the senate. The electoral connection in Argentina works quite differently than in the United States. In Argentina, electoral mechanisms make provincial governors (as provincial party leaders, with substantial control over party list formation) powerful actors in national politics, rather than individual legislators. At the same time, although provinces have large spending responsibilities, they raise little in taxes. Most of the provinces’ funding comes from a common pool of resources collected by the national government for itself and for the provinces. These two features together—control by governors over their legislators, and the importance of the national government in provincial public finances—are the backbone of a symbiotic crisscrossing of national and provincial politics and policies, which is described in more detail later. Argentina’s first constitutional president took office in 1862. The formal machinery of democracy, elections, and checks and balances operated in Argentina until 1930, the first time that a military coup succeeded in removing a constitutionally elected president. Between 1930 and 1982, twelve presidents (both de jure and de facto) were removed from office by force. Since the 1940s, the political scene has been dominated by two parties, the Peronist Party (Partido Justicialista, or PJ) and the Unión Cívica Radical (UCR). The UCR, the oldest active party, emerged during the 1890s to challenge oligarchic rule. It first captured the presidency in 1916, after the democratizing electoral reforms of 1912. The PJ emerged in the 1940s under the charismatic leadership of Juan Domingo Perón (president from 1946 to 1955, and from 1973 until his death in 1974) and has since become the main political party in Argentina. Historically, the socioeconomic base of UCR has been the middle class, and the socioeconomic base of PJ has been the working class. Even though those associations have remained accurate to some extent until the present, both are considered catchall parties whose positions nowadays revolve around the political “center.” Peronism has traditionally combined a metropolitan coalition based on the urban working class with a periph-

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

eral coalition of provincial groups from the country’s more “backward” interior (Gibson 1997; Gibson and Calvo 2000). Even though a majority of the voters in these peripheral provinces are poor, provincial PJ party machines tend to be dominated by political elites with ties to provincial economic power. While third parties (of the left and right) have on occasion achieved some political prominence, to date all of these parties have had an ephemeral existence, rarely lasting more than five years as a relevant political force. This failure is due primarily to the parties’ inability to build a territorial reach like that of the Peronist and (secondarily) Radical parties in a country where national politics have a strong “subnational” drag (Torre 2003; Calvo and Murillo 2004). Recent events, like the disastrous ending of the De la Rúa government, have weakened the UCR. The subnational drag on national politics described below and the overrepresentation of small backward provinces give inordinate weight in Argentine politics and policymaking to the clientelistic backwaters of the country, and underrepresents the more modern, urban, ideological, and programmatic segments of Argentine society. Congress, Political Careers, and the Provincial Connection

Legislators in the United States enjoy long careers in congress and have high reelection rates, they specialize in powerful committees, and they tend to be high-quality politicians; the U.S. Congress is a central arena of policymaking. In Argentina, congress has exactly the opposite characteristics. Argentine legislators have very low reelection rates and tend to serve short terms in congress; congress is not a very important policymaking arena; it is not very institutionalized, and by and large, it is not the place where policy expertise resides. The assessment of legislator quality in the United States is based on the extremely careful work of Diermeier, Keane, and Merlo (2005), who, using labor econometrics techniques, estimate the unobserved characteristics of U.S. legislators from 1947 to 1994. We have no way of replicating such detailed work at this point, but Stein and Tommasi (2005a) have constructed a preliminary indicator of the policymaking capabilities of

Pablo T. Spiller

and

Mariano Tommasi

18 Latin American congresses. That index, built from input from Saiegh (2005) and references there, attempts to capture some characteristics of each legislature as an organization and of its legislators, which relate to the capacity to engage in substantive policymaking activity. Table 3.1 reproduces that index for the Latin American sample; Argentina is one of the five countries with the lowest value of the index. The table highlights, in a comparative perspective, several points we want to make about the Argentine congress and about Argentine legislators, including the fact that congress is a weak institution in terms of legislative capabilities, the fact that legislators do not stay long in congress, and the fact that congress is not as important for political careers in Argentina as it is in some other countries. We address these issues in turn below. Legislative Organization

As in most legislative bodies across the world, congressional committees in Argentina are the workhorses of the legislative process. Yet in the Argentine case, the policy jurisdictions of congressional committees do not parallel the structure of administrative or cabinet agencies. Some committees have far too vast (and others far too narrow) a focus. Not only are legislative committees poorly matched with the structure of the cabinet, but their number and size also bear no correspondence to the size of the legislature. Furthermore, even though from 1983 to 2004 the number of deputies increased only slightly (from 254 to 257), the number of standing committees increased from 27 to 45. Danesi (2004) argues that the creation of these new committees had more to do with the need to assign a committee chairmanship or other position to some important politician than to legislative needs. (This is reflected in the vague wording used to justify the creation of each new committee.) Not only has the number of committee slots increased monotonically, but so has the tendency of legislators to belong to them. In 1997 the typical Argentine deputy served on 3.5 committees, but by 2004 that deputy served on 4.5 committees: a much higher number than in neighboring countries like Chile or Brazil (Danesi 2004). This overstretching of legislators across

20.5 19.9 24.9 36.0 20.3 29.9 n.a. 13.3 27.7 19.9 30.8 27.4 23.1 22.5 25.0 22.1 38.2 27.8

1.6 1.8 3.1 3.7 2.7 2.2 2.0 1.7 2.1 1.8 2.6 2.0 1.6 1.8 2.2 1.7 2.7 1.4

2.9 3.3 5.5 8.0 4.0 2.6 3.1 3.5 3.9 3.2 3.0 1.9 3.5 5.8 5.5 5.2 8.8 4.9

69.6 78.4 54.0 79.4 91.6 8.4 49.6 83.1 64.0 68.4 73.1 89.5 85.6 81.3 75.4 92.9 68.4 74.6

4.50 1.66 0.92 1.95 0.86 2.09 3.54 1.26 2.44 3.24 2.34 2.43 1.96 1.86 3.15 2.44 0.98 0.98

Medium Medium Medium High High High Low High Medium Low Low High Low Medium Low Low High Medium

Average number of committee memberships Strength of per legislator committees Low Medium High High High Medium High Medium High Medium Low Medium Medium High High Low High Medium

Low Medium High High Medium Low Low Low Low Low Low Medium Medium Low Low Low Low Low

Place to build Technical career expertise

Low Medium High High High Medium Low Medium Medium Low Low Medium Medium Medium Medium Low High Medium

Congress Capability Index

and

Policy Outcomes in

Source: Stein and Tommasi (2005a) and references there. n.a.= not available

Argentina Bolivia Brazil Chile Colombia Costa Rica Dominican Rep. Ecuador El Salvador Guatemala Honduras Mexico Nicaragua Panama Paraguay Peru Uruguay Venezuela

Percentage of Confidence Average legislators in congress, Effectiveness experience of with average of law-making legislators university Country 1996â&#x20AC;&#x201C;2004 bodies (years) education

Table 3.1. Policymaking Capabilities of Latin American Congresses

Political Institutions, Policymaking Processes, Argentina

Pablo T. Spiller

and

Mariano Tommasi

committees leads to little focus and specialization, especially given the brevity of legislative careers (as explained below). Jones, Saiegh, Spiller, and Tommasi (2002) provide an empirical analysis of the composition of legislative committees in Argentina. Our objective was to test whether some of the theories of legislative organization developed for the U.S. case applied to the Argentine case. We rejected the so-called “distributive hypothesis,” according to which legislatures are organized so that legislators can obtain special decision power in those policy areas of more interest to their voters (Weingast and Marshall 1988). Constituency interests, although relevant to committee membership, are politically insignificant. We also evaluated the “informational hypothesis,” according to which legislators are sorted into committees in order to foster specialization and investment in knowledge necessary to deal with complex technical issues (Krehbiel 1991). A casual review of the empirical results in Jones, Saiegh, Spiller, and Tommasi (2002) may appear to provide support for that hypothesis, as the professional background of legislators is significant in the allocation of committee assignments. A closer examination of the results, however, leads to the conclusion that support for the informational hypothesis is also modest. Even though background may determine committee membership, short tenure and multiple memberships weaken any notion of specialization. As a consequence, informational advantages are not obtained. Additional knowledge of the Argentine legislature gained from extensive interviews with legislators, staffers, and other qualified observers supports this view (Danesi 2004). A more compelling explanation of these results is that legislators, required to serve on committees, simply choose those whose topic they find of greatest intrinsic interest and on which they can serve at the least personal cost. The picture is slightly different when one analyzes committee chairmanships, as opposed to just committee membership. In contrast to serving on a committee as a general member, chairing a committee is highly valued. Most committee chairs receive extra resources, mostly for staff salaries, amounting to approximately 50 percent of the base allocation received by each legislator. The committee chair also controls the permanent staff assigned to the committee: a secretary, an administra-

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

tive secretary, and two clerical assistants. These personnel perform only administrative functions, and should be seen as a way to compensate loyal legislators with patronage positions rather than as an indication of a chair’s predisposition to conduct a committee’s business in a professional manner. Committee chairmanships are more coveted positions given this access to greater resources. In our analysis we found that it is often the case that these positions are occupied by slightly more senior legislators with good relations with their provincial governors, and that these legislators tend to stay in such chairs a little longer than regular members do in committees (Jones, Saiegh, Spiller, and Tommasi 2002). This evidence indicates that the Argentine legislature is not organized in a manner that maximizes its effectiveness in the policymaking process. Additional evidence of Argentine deputies’ low level of interest in legislation is reflected in the allocation of resources within the chamber. Danesi (2004) presents evidence comparing the Argentine chamber of deputies with those of other Latin American countries (Chile, Paraguay, and Uruguay). Argentina ranks lowest in several indicators of the resources devoted to legislative functions. Argentine legislators are simply not too worried about making the Argentine congress into a strong and capable policymaking body. This is an outcome of their career incentives, which in turn relates to the brevity of congressional careers, a point to which we turn next. The Brevity of Legislative Careers

The first thing to note about the careers of Argentine legislators is that they are quite short. As of January 2001, only one legislator had served continuously in the chamber of deputies since the country’s return to democracy in 1983. Since 1983, the overall stability of membership in the Argentine chamber of deputies has been relatively low. During the 1983–2003 period, the typical Argentine deputy served only one term in office, and only 20 percent of incumbents were reelected to their seats (see table 3.2). In contrast, during the twentieth century, the average U.S. House member served between five and six terms (Ornstein, Mann, and Malbin 1998).

Pablo T. Spiller

and

Mariano Tommasi

Table 3.2. Reelection Rates to the Argentine Chamber of Deputies, 1985–2003 Year 1985–87 1987–89 1989–91 1991–93 1993–95

Percentage of deputies reelected Year 29.2 22.0 18.9 16.2 14.2

Percentage of deputies reelected

1995–97 1997–99 1999–2001 2001–2003 1985–2003 average

14.7 20.4 23.6 15.4 19.4

Source: Authors’ elaboration with data from Molinelli, Palanza, and Sin (1999) and from official records of the Argentine Chamber of Deputies.

The distribution of legislators’ number of terms served in the Argentine congress is almost identical to that of Costa Rica, a country with term limits where legislators are forbidden from running for consecutive reelection. This can be seen in table 3.3, which shows the number of terms served by legislators in the Costa Rican Assembly from 1949 to 1990 alongside the same information for Argentine legislators from 1983 to 2001. Reelection rates are quite low in comparative perspective. Table 3.4 shows reelection rates for a number of countries. Argentina presents the lowest reelection rate except for Mexico, where reelection is not allowed. Why are Argentine legislators not reelected? One possible reason is that the voters are throwing the rascals out. But a closer look at table 3.4 shows that the low reelection rate is not always the voters’ choice: only Table 3.3. Congressional Service by Argentine and Costa Rican Legislators Number of terms served 1 2 3+

Number of legislators Costa Rica (1949–90) Argentina (1983–2001) 87 11 3

85 11 4

Source: Carey (1996) for Costa Rica, and authors’ elaboration from information in Molinelli, Palanza, and Sin (1999) and in official records of the Argentine Chamber of Deputies for Argentina.

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

Table 3.4. Reelection Rates in Selected Countries, Various Years Percent seeking Country reelection Argentina (1997) Brazil (1995) Chile (1993) Japan (1963–90) Mexico (1997) United States (1996) Colombia (1990) United Kingdom (1950–74) Italy (1953–72) Panama (1999) Portugal (1991) Turkey (1950–80) West Germany (1957–76)

Percentage of candidates reelected

Total percent reelected

26 70 76 91 0 88

67 62 78 82 0 94

17 43 59 74 0 83 48 81

82 49 57.8 56 70–75

Source: Morgenstern (1998); Archer and Shugart (1997); Molinelli, Palanza, and Sin (1999). Note: Empty cells reflect that figures are not available.

a small percentage of legislators seek reelection. Of those who sought reelection in Argentina, 67 percent succeeded: a number that is not particularly low in international comparison. But only 26 percent actually sought reelection. Having established that the high rotation of Argentine legislators is not due to voter choice, in the next couple of sections we study in more detail the nature of Argentine political careers, as well as how the decision is made whether or not to renominate legislators to congress. Political Careers

Members of the Argentine congress do not develop long “professionalized” legislative careers. Yet, they are not political outsiders either. That led us to describe them as “amateur legislators but professional politicians” in Jones, Saiegh, Spiller, and Tommasi (2002). Even though most deputies and senators serve only one term in congress, almost all of them occupy some government or party position before and after their

Pablo T. Spiller

and

Mariano Tommasi

congressional stint. These positions could be national, provincial, or municipal, but in most cases are tied to politics at the provincial level. Jones (2004) provides information on the positions held by Argentine legislators before their congressional terms. He studies the prior positions held by deputies elected between 1991 and 1999 and senators elected between 1986 and 2001. One of the main conclusions stemming from his analysis is that virtually all deputies (97 percent) and all senators occupied either a governmental or party position immediately before their election. This fact underscores the presence of relatively stable career pathways in Argentine political parties. Another important fact that comes out of that analysis is that a majority of those governmental or party positions were in the home province of deputies (62 percent) and senators (54 percent). The province-centeredness of political careers is stronger among legislators from the dominant PJ and UCR parties (and obviously from provincial parties) than among the scarcely represented and often short-lived other national parties, such as FREPASO. Jones, Saiegh, Spiller, and Tommasi (2003) performed an analysis of the position held by major party (PJ and UCR) deputies of the 1991â&#x20AC;&#x201C;95 class as of mid-1988, two and a half years following the end of their term in office. That analysis also highlighted that a large percentage (85 percent) of deputies continued to occupy a partisan or governmental post, and that many of those posts were province-based. Nearly 60 percent of the deputies who had left congress held a political position at the provincial level. The next section looks at the provincial connection in more detail. The Subnational Connection: The Role of Provincial Party Bosses

As the above discussion suggests, the keys to career advancement are held by the provincial leaders of the parties to which legislators belong. Therefore, it is worth taking a closer look at these leaders, who we call â&#x20AC;&#x153;provincial party bosses.â&#x20AC;? That task is undertaken in Ardanaz, Leiras, and Tommasi (2007), studying who these bosses are, identifying their sources

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

of power, and showing how they interact in national policymaking, especially through congress and in key policy events. The important role played by party bosses is suggestive of how subnational (provincial-level) politics influences national politics. The 23 provinces plus the city of Buenos Aires serve as electoral constituencies for congressional elections. This makes the province the locus of party competition and the base of political support for politicians and parties.4 As shown above, moreover, political careers are usually province-based, and even positions in the national government are often a consequence of provincial factors. It is common for a single person or small group of politicians to dominate political parties at the provincial level. In provinces where the party controls the governorship, the governor is, with rare exceptions, the undisputed (or at least dominant) boss of the provincial-level party. In many other provinces where the party does not hold the governorship, the party is nonetheless dominated in a comparable manner by a single individual, but there is a greater amount of space for intraparty opponents. In the remaining provinces where the party does not control the governorship and there is not a single dominant leader, there is usually a small group of influential party leaders who predominate in party life. Following the office of the president, the governorship is the most important institutional position in the Argentine political system. In fact, the governorsâ&#x20AC;&#x201D;generally in a collective mannerâ&#x20AC;&#x201D;have constituted a more relevant counterweight to the presidential authority than congress or the judiciary (De Luca 2004). The two major parties have exerted a great dominance at the subnational level. The PJ controlled an average of 61.8 percent of the governorships between 1983 and 2003 (ranging from 54.6 percent to 77.3 percent), with the UCR placing second with an average of 23.3 percent (ranging from 9.1 percent to 33.3 percent). No other party ever possessed more than one governorship at any one time during this period. This fact has been highlighted by a spurt of recent literature, which is reviewed in Ardanaz, Leiras, and Tommasi (2007). 4

Pablo T. Spiller

and

Mariano Tommasi

Another noticeable feature of provincial governorships is the high degree of name repetition (Ardanaz, Leiras, and Tommasi 2007). This is an indication of the fact that many provinces are controlled by single individuals or families over extensive periods of time.5 Finally, a review of this name list shows that some of the most salient national political figures have been long-time provincial governors. In fact, all presidential candidates from the major parties (the PJ and UCR) were governors before running in presidential races.6 The power of provincial party bosses has been reinforced over time since the return of democracy by two channels. On the one hand, there have been political reforms, induced by these very leaders, which have increased their power within the provinces. For instance, no provincial constitution allowed governors to be reelected in 1983; by 2003 all but 6 of the 24 provinces had provided for the immediate reelection of the governor, 4 of them without imposing restrictions on the number of terms that a governor could serve.7 Incumbent governors have also manipulated other institutional levers at their disposal in order to increase their control of the provincial political arena, including the timing of local elections vis-à-vis national ones. Electoral rules allow provincial

Gibson (2004) provides a vivid account of one of the strongest examples of what he terms “subnational authoritarianism” in the province of Santiago del Estero, which has been “owned” by the Peronist caudillo Carlos Juárez since he first assumed the governorship in 1949. Bill Chavez (2003) provides a somewhat similar depiction for the province of San Luis, which has been dominated for many years by Adolfo Rodríguez Saá (briefly Argentine president during the December 2001 crisis) and his family. See also accounts by local political anthropologists about Misiones and San Luis, summarized in Cleary and Stokes (2006). 6 More generally, some of the most salient national political figures since the return to democracy have been long-time provincial governors or members of family-run provincial governments. These include the cases of Angeloz (UCR, Córdoba, 1983–95), Duhalde (PJ, Buenos Aires, 1991–99), Kirchner (PJ, Santa Cruz, 1991–2003), Massaccesi (UCR, Río Negro, 1987–95), Menem (PJ, La Rioja, 1983–89), Rodríguez Saá (PJ, San Luis, 1983–2001), and Ramón Saadi and Vicente Saadi (PJ, Catamarca, 1983–91). 7 Three of those four provinces with indefinite reelection are the provinces where the three Peronist presidential candidates in the 2003 election hail from—La Rioja, San Luis, and Santa Cruz. That is another indicator of the connection between subnational bailiwicks and national political power. See Calvo and Micozzi (2005) and Ardanaz, Leiras, and Tommasi (2007). 5

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

governments to set the date of local elections. Governors have exploited their control over local legislatures and manipulated the electoral calendar in order to avoid negative (coattail) externalities from the national arena and in order to take advantage of positive coattail externalities from the popularity of national presidents (Calvo and Micozzi 2005; Ardanaz, Leiras, and Tommasi 2007). Additionally, the market-oriented reforms of the 1990s have increased the power of governors in at least two dimensions.8 On the one hand, the decentralization of spending has transferred more fiscal (and hence political) resources to the provincial governments. On the other hand, welfare strategies have been transformed from national labor protection policies to compensatory social programs, programs that, even though national, are administered politically and clientelistically by provincial machineries (Weitz-Shapiro 2007; Levitsky 2007). In sum, the dominance exercised by provincial party leaders over the political process in their respective provinces is based principally on patronage, pork barrel politics, and clientelism (Calvo and Murillo 2004). Campaigns, both primary and general election, are funded primarily through the use of resources gained from patronage, pork barrel, and clientelistic activities. Government financing of campaigns and party building also exists, although it represents only a very modest fraction of the resources used for campaign activity by the relevant parties (Jones 2004). Patronage positions are particularly important for maintaining the support of second- and third-tier party leaders, who in turn possess the ability to mobilize voters, especially for party primaries. The ability to engage in pork barrel politics improves the partyâ&#x20AC;&#x2122;s reputation with key constituents and aids clientelistic practices through the provision of jobs to party supporters and the infusion of money into the party coffers, which in turn is employed to maintain clientelistic networks. Clientelism assists party leaders at all levels in maintaining a solid base of supporters. Provincial party bosses are able to dominate local politics by mobilizing a variety of resources. Their powers are not restricted to the confines 8

Ardanaz, Leiras, and Tommasi (2007) explore these two dynamics in more detail.

Pablo T. Spiller

and

Mariano Tommasi

of their respective provinces. There are political and institutional variables that enable governors, through their control of legislative contingents in the national congress, to enjoy national leverage. One crucial such variable is the control of candidate selection methods. Candidate Selection

Political parties, not the government, run party primaries for both party leadership positions and candidacies for national, provincial, and municipal public office.9 Primaries involve a considerable amount of votermobilization efforts on the part of the competing intraparty lists. The electorate for these contests consists of either party members alone or party members and those unaffiliated with any party. When a primary is held, success thus depends almost entirely on candidatesâ&#x20AC;&#x2122; financial or material resources. Whether or not a politician will faithfully represent (or has faithfully represented) the interests of his or her constituents normally has little impact on success in the primary contest. It is in the best interest of the regional party boss to avoid internal conflict, since conflict increases the influence of the national party and the independence of its legislators. Because governors are likely to bear a disproportionate share of the costs of any divisive primary, they have both the incentive and the means to arrange a negotiated list of candidates. Most governors are able to impose their candidates, co-opt potential opponents, or successfully negotiate an agreement with other party factions. Given the high reelection rate of governors who seek continued office, the power of the governor depends in part on whether the provincial constitution restricts reelection. In contrast, where the provincial-level party is in opposition at the provincial level, the resources at the disposal of its leader are minimal in comparison to those held by a governor; its leader is therefore much less likely to be undisputed. In such circumstances, national-level organizations have more influence on the provincial list composition, and the chance that a primary will be held is higher. 9

This subsection draws from De Luca, Jones, and Tula (2002) and Jones (2004).

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

The decision within the PJ and UCR to hold a primary depends first and foremost on whether the party controls the governorship at the provincial level and secondarily on whether the incumbent governor is eligible to seek reelection. Regardless of which method is used, the provincial-level party leaders are the key players in the nomination process, with the national party leadership and rank-and-file members playing a decidedly secondary role. Furthermore, even when a direct primary election is held, the weight of the party leadership in the outcome of the contest is quite powerful, since unlike in current U.S. primaries, Argentine primary elections are clashes of party machines, with the victory going to the machine that is able to mobilize the largest number of voters. As JosĂŠ Luis Lizurume, Chubut governor from 1999 to 2003, recently stated, â&#x20AC;&#x153;La interna es aparato puroâ&#x20AC;? (The primary is just machine).10 The importance of province-level political variables in explaining the method used to select candidates for the national congress underscores the decentralized nature of the Argentine party system and highlights the prominent influence of provincial politics on national politics. The Argentine Senate in Comparative Perspective

To give some comparative flavor to several of the points raised so far, we close this section with a brief comparison of the Argentine and Brazilian senates. We compare Argentina to another strongly federal country: a case which has been characterized as one in which provincial governors are very important in the national political arena, and in which political careers are supposed to center strongly in the subnational units (the Brazilian states) (Samuels 2000a, 2000b, 2003). This discussion is based on a survey conducted and evaluated by Mariana Llanos (2003b). That survey of Argentine and Brazilian senators suggests that the Argentine senate is a weaker policymaking body with lower technical capabilities than its Brazilian counterpart. Committees in the Argentine senate are perceived to be less important than in Brazil. Argentine senators seem to look for instructions from provincial 10

Jones (2004), citing the July 18, 2003 edition of Diario El Chubut.

Pablo T. Spiller

and

Mariano Tommasi

governments to a much greater extent than do their Brazilian peers. They see their main job as obtaining resources for the province. When conflicts arise, they vote more according to the needs of their province and less according to the party line than do Brazilian senators. These are remarkable findings given the presumption in the specialized literature about the strong subnational drag to legislative careers and policymaking in Brazil (see, for instance, Samuels 2000b and 2003).11 Some of the evidence in Llanos’s study is summarized below. Regarding the importance of the subnational connection, 85 percent of the surveyed Argentine senators expressed the view that the opinion of their provincial government is very important in their decisions, compared to just 64 percent of the senators surveyed in Brazil. Two-thirds of Argentine senators stated that obtaining resources for their respective provinces is a very important part of their legislative activity. Only 45 percent of the senators surveyed in Brazil shared this opinion. When asked whether they would side with their provincial interests or with their party in case there was a conflict of interest, 80 percent of Argentine senators said they would always favor their province’s point of view, and only 6 percent said they would always vote with the party. In the case of Brazil, only 55 percent of the senators said they would always vote with their province, while 13 percent answered that they would always favor their party’s point of view. Regarding the importance of legislative committees, most of the Argentine senators expressed a negative opinion about the workings of the committee system in the Argentine senate. Nine out of ten respondents said that the functioning of the system was hindered by the existence of too many committees. In Brazil, only 21 percent of the senators offered a similar response. Also, more than half the respondents in Argentina said that committees are not an arena for the technical discussion of bills. Only 11 percent of the Brazilian senators expressed a similar response. That important interpretation of Brazilian politics is somewhat challenged in chapter 4 in this volume by Lee Alston, Marcus André Melo, Bernardo Mueller, and Carlos Pereira, who claim that the subnational dominance of Brazilian politics has been dwindling over time. Even in that more nuanced account, Brazil still appears as a country in which subnational politics is an important component of the overall system. 11

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

Finally, 87 percent of the senators in Argentina said that the work of committees is severely underappreciated by their peers, who seldom participate in the meetings. Half the senators in Brazil shared this opinion. In terms of legislative oversight, 70 percent of the Argentine senators expressed the opinion that investigative committees do not play an important role in making governmental activities more transparent. This opinion contrasts sharply with that of their Brazilian peers: 98 percent of them expressed the opposite opinion. Regarding legislative work, floor debates seem to play a more important role in the voting decisions of Brazilian senators. Thirty-three percent of the Brazilian respondents, against only 9 percent of Argentine respondents, said that they took floor debates into consideration when deciding whether to support a bill. In every single category, the views of Brazilian senators with regard to the technical inputs at their disposal were significantly more favorable than those of Argentine senators. The inputs in question included technical assistance, the information office, computer services, the library, and infrastructure and equipment. Summarizing, this section applies the insight that legislative behavior and the organization of legislative institutions are affected by electoral rules to reveal some basic features of Argentinaâ&#x20AC;&#x2122;s national legislature. By making legislators more beholden to the provincial party boss than to the voters, Argentinaâ&#x20AC;&#x2122;s electoral rules, along with a constitutional system that places limited constraints on unilateral executive actions, have created an amateur congress, one whose members have neither the expertise nor the incentives to initiate influential legislation, to control public administration, or even to invest in strengthening congressional institutions. These characteristics of the Argentine congress, in turn, have important implications for the qualities of public policy in Argentina. The evidence presented here shows, though, that these legislators are not amateur politicians. Argentine legislatorsâ&#x20AC;&#x2122; progressive ambition causes them to leave congress, but not politics. The center of political careers is in the provinces. In turn, provincial politics is heavily influenced by the objectives and resources of provincial governors. The next section focuses on the interactions of provincial governors and national political actors in the crucial domain of fiscal federalism.

Pablo T. Spiller

and

Mariano Tommasi

The Federal Fiscal Game

In Argentina, national and subnational politics and policies are intertwined to a much larger (and convoluted) extent than in other federal polities.12 The main links are electoral and fiscal. Provinces undertake a large fraction of total spending, yet collect only a small fraction of taxes. Provincial spending amounts to 50 percent of total consolidated public sector spending. This figure rises to close to 70 percent if one excludes the pension system and focuses on “more discretionary” spending. Furthermore, the type of spending in the hands of provincial governments tends to be politically attractive (such as public employment and social programs) because it is close to the interests of territorially based constituencies. Yet on average, provinces finance only 35 percent of that spending with their own revenues. The rest of their spending is financed from a common pool of resources, according to the “Federal Tax-Sharing Agreement.” In a large number of small provinces, the proportion of funds from this common pool constitutes over 80 percent of their funding. Local politicians, then, enjoy a large share of the political benefit of spending and pay only a small fraction of the political cost of taxation. This fiscal structure at the provincial level is one reason why many professional politicians are more interested in pursuing a career through appointments in the provincial government (or even the party at the provincial level) than in the national congress. But the powerful provincial brokers—that is, the governors—depend heavily on the allocation of “central” monies to their provinces to run both their political and their policy businesses. That is, they need central money to deliver particularistic political goods, as well as to provide general public goods in their province. There are several channels for funneling funds to the provinces; the main ones are the geographic allocation of the national budget and the Federal Tax-Sharing Agreement.

This section is based on Saiegh, Spiller, and Tommasi (2007) and Tommasi (2006), and references there. 12

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

The game in which these allocations are determined is the source of many political and policy distortions, at both the national and the provincial levels. The game even affects the quality of democracy at the local level. The Argentine voter at the provincial level has an incentive to reward politicians who are effective at extracting resources from the center.13 These are not necessarily the most competent or honest administrators.14 Given the political mechanisms by which funds are allocated, this also adds uncertainty to provincial public finances, since it is not easy to project future allocations, and lowers the quality of provincial policies. The history and evolution of the Tax-Sharing Agreement is fraught with examples of opportunistic manipulation, occasionally curtailed by fairly rigid and inefficient mechanisms (Tommasi 2006; Iaryczower, Saiegh, and Tommasi 1999). Unilateral, bilateral, and coalitional opportunism (by the national government, by a province, or by a set of provinces that turns out to be pivotal for an important vote in congress or for some other reason) has been common in the allocation of central government monies to the provinces. The national executive has enjoyed substantial discretion to allocate items in the federal budget geographically (Bercoff and Meloni 2007). In an attempt to prevent adverse changes in the future (for instance, a reduction in the amounts going to any specific province), political actors have tended to impose greater rigidity on the Tax-Sharing Agreement, reducing the government’s capacity to adjust fiscal policy to changed economic circumstances. One example is the earmarking of taxes for specific programs with clear regional distributional effects. This earmarking led to a rigid and convoluted system of federal tax collection and distribution, which has been christened the “Argentine fiscal labyrinth.” Recent attempts to simplify that labyrinth, which also reflect the inability to strike efficient intertemporal agreements, led to the 1999 and 2000 “fiscal pacts” between the national and provincial governments. Jones, Meloni, and Tommasi (2007) show that voters in Argentina, unlike those in the United States, reward provincial spending in gubernatorial elections. 14 On the heterogeneity of local democracy in Argentina, and its connection to fiscal federalism, see Sawers (1996), Cleary and Stokes (2006), and references there. 13

100

Pablo T. Spiller

and

Mariano Tommasi

Those pacts generated a rigid commitment to a minimum of revenues from the center to the provinces, which turned out to be a very costly straightjacket for the De la Rúa government during the lead-up to the 2001 crisis. Similarly, the lack of cooperation from the provinces has been credited as the immediate cause of the country’s move to default (Eaton 2005; Tommasi 2006). That episode of the Argentine federal fiscal drama, which led to one of the largest defaults in modern world economic history, was a clear manifestation of one of the central points in our argument. Provincial governors, who are crucial players in national politics and policymaking, and who might have a slightly longer horizon than other players, have only secondary interest in national public goods (such as macroeconomic stability), and in the quality of national policies, and (hence) in investing in institutions (a professional congress, a stronger civil service) that might improve the quality of policies. Their primary interest—on the basis of which they grant or withdraw support to national governments and their policies—is the access to common-pool fiscal resources. Weak Judicial Enforcement

The workings of judicial institutions have direct implications for the feasibility of private contracting, contracting among private and public actors, and arrangements among political agents. This section focuses on the latter, emphasizing the role of the supreme court as potential enforcer of constitutional and legislative contracts. Over the past several decades, the Argentine supreme court has not been a strong and impartial enforcer of political agreements. The reasons for weak judicial enforcement lie more in politics than in a lack of jurisprudence.15 Iaryczower, Spiller, and Tommasi (2002, 2007) show that a strategic behavioral model similar to the one used to explain the U.S. supreme court explains quite well the behavior of Argentine supreme court justices. Furthermore, the fragmentation of the Argentine polity would suggest that the Argentine supreme court should be strong and independent. 15

This section is based on Iaryczower, Spiller, and Tommasi (2002, 2006, 2007).

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

But Argentina’s political history has resulted in a court with a much smaller role than the supreme court has in the United States. In particular, since the mid-1940s, Argentine supreme court justices have had very short tenures—indeed, among the shortest in the world. From 1960 until the mid-1990s, the average Argentine justice lasted less than five years in his post. This short average tenure puts the country near the bottom of the ranking, alongside countries not usually associated with stability and the predominance of the rule of law (such as Malawi, Pakistan, and Peru), and a long distance from the United States, where supreme court justices serve an average of around 20 years, or Norway and New Zealand, where average tenure is longer than 15 years (see table 3.5). Table 3.5. Supreme Court Justices’ Average Tenure, Selected Countries, 1960–1990 (number of years) Country USA Norway New Zealand Malaysia Ireland Australia Singapore Canada Belgium Guyana Germany, FR Netherlands Italy Jamaica Trinidad&Tobago Nigeria South Africa Bangladesh Brazil Nicaragua Sudan France Israel Source: Henisz (2000).

Judicial tenure 18.8 16.2 15.3 14.7 14.7 14.6 14.5 14.2 13.2 12.7 12.2 12.1 11.1 10.9 10.6 10.1 8.3 8.1 7.2 7.1 6.8 6.6 6.5

Country United Kingdom Sri Lanka Chile Zambia Ghana India Botswana Philippines Zimbabwe Kenya Malawi Pakistan Argentina Peru Dominican Rep. Cameroon Mexico Honduras Colombia Ecuador Guatemala Paraguay

Judicial tenure 6.4 5.8 5.7 5.6 5.5 5.4 5.2 4.9 4.6 4.6 4.6 4.4 4.4 4.0 3.6 3.3 3.3 2.8 1.9 1.9 1.8 1.1

101

102

Pablo T. Spiller

and

Mariano Tommasi

Short judicial tenure is a feature of the past 50 years. Indeed, after World War I the Argentine court was on a path of convergence with the U.S. supreme court. From the creation of the Argentine supreme court in 1863, the average tenure of justices increased systematically, reaching the same level as in the United States during the 1920s. Then during his first administration in 1946, President Perón impeached the sitting supreme court justices: an act that had a lasting impact. From that point on, the norm of not manipulating supreme court membership weakened substantially. The several military and civilian presidents who alternated in power appointed their own courts. In 1991, the first time since 1946 that a president might have faced an opposition court, President Carlos Menem expanded the size of the court from five to nine members, thereby granting him a “working” judicial majority. Indeed, control over the court was such that, since the mid-1940s until De la Rúa’s inauguration in 1999, no president faced a court with a majority appointed by a political adversary. Similarly, since the mid-1940s until the present, no president, except for De la Rúa, has left the supreme court intact. All others have directly manipulated its composition, whether by removing justices (either directly or through threats of impeachment), or by expanding or shrinking its composition. Courts with short judicial tenure naturally tend to be aligned with the sitting government, and hence are unlikely to strongly wield their power of judicial review. During a large part of the twentieth century—due mostly to de facto governments, but also to the political alignment during the interim democratic spells—executives often enjoyed a high level of political support in the legislature. This alignment is also a variable that leads the supreme court not to challenge the government. The dynamics of a court without much clout are perverse. Lack of clout and loss of public respect are self-reinforcing. Indeed, during recent presidential elections, several presidential candidates promised to remove (some or all of) the sitting justices, and current president Néstor Kirchner has managed to alter the composition of the court in his favor.16 Miller (1997) provides an interesting sociological interpretation of the historical weakening of the Argentine supreme court by comparing its trajectory to that of its U.S. counterpart. 16

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

A Bureaucracy without a Long-Term Principal

One possible way to enforce intertemporal political agreements is to delegate enforcement to a relatively independent, yet accountable bureaucracy. Argentina, however, does not have such a bureaucracy.17 Even though Argentina has a more developed civil service system than some of the poorer Latin American countries, political shortsightedness and lack of consistency have contributed to weaken its bureaucratic apparatus way below what one could expect from a country with the level of human capital of Argentina. Figure 3.2 plots an index of â&#x20AC;&#x153;Weberiannessâ&#x20AC;? of the bureaucracy (from Rauch and Evans 2000) against an index of human development from the United Nations. Argentina appears as an outlier with a very weak bureaucracy compared to its level of development. Even though from time to time there have been islands of bureaucratic excellence in the Argentine public administration, often those islands have not been respected when such independence collided with the short-term political interest of the government of the day. For instance, the current administration has not been keen on the independence of the central bank or of the National Statistics Office (INDEC), but has strengthened the Tax Collection Agency. Bambaci, Spiller, and Tommasi (2007) provide further evidence of the stop-and-go nature of capacity building in the Argentine public sector. The absence of long-term principals hinders the building of a professional bureaucracy. Executives in almost all presidential systems are transient; but in Argentina, where the members of congress also are not long-term principals, they are not motivated to control the administration either. As Krehbiel (1991) argues, legislators everywhere tend to undersupply public goods such as controlling the administration. That effect is magnified in Argentina, where legislatorsâ&#x20AC;&#x2122; key incentive is to attend to the interests of provincial party leaders who, in turn, are not particularly interested in the quality (or in most cases even the content) of national policymaking. 17

This section is based on Bambaci, Spiller, and Tommasi (2007).

103

Pablo T. Spiller

and

Mariano Tommasi

Figure 3.2. â&#x20AC;&#x153;Weberiannessâ&#x20AC;? of the Bureaucracy in Relation to Human Development in 35 Developing Countries 14

Sgp Kor

12 Pak

Mys

Ind

Webscale

104

Tun Egy

Civ

Mar

Tha

L ka Tur

Bra

Isr Chl

Per

Nga 2

Spn

Cri Mex

Col

Phl Prt

Ury

Ecu Syr

Hti

Hkg

Grc

Arg

Gtm Dom

Ken 0 0

100

120

140

160

180

UNDP Human Development Index 2004 Rank: 1=Lowest development 177=Highest development Source: Rauch and Evans (2000) and UNDP Note: Latin American countries are highlighted.

When the bureaucracy, as a consequence, faces weak long-term incentives, its members tend to ignore their duties, and intrusive administrative controls to prevent corruption must be put in place, which further reduce the ability to generate timely and effective policies. Each new executive, unable to motivate (or fire) the permanent bureaucracy, nominates large numbers of additional political appointees, usually under much more flexible labor agreements; they can easily be fired, which they very often are after a change in their direct overseer. These practices have created a transient parallel bureaucracy. The parallel bureaucracy undertakes the same actions that the normal bureaucracy is designed to effect, but is unable (or unwilling) to undertake. The turnover at the ministerial and secretarial levels also implies turnover in the parallel bureaucracy. This turnover limits the extent of institutional knowledge and the development of cooperation across ministries and secretariats,

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

deepens the heterogeneity in policy quality, and reinforces the lack of policy coherence. It is well known that the parallel bureaucracy in Argentina is widespread, but its extent is difficult to measure. In Bambaci, Spiller, and Tommasi (2007), we present some evidence that the parallel bureaucracy constitutes a large fraction of the top echelons of the public sector. The members of the parallel bureaucracy tend to be better paid, but shorter-lived than permanent civil servants at similar levels. They are hired under a variety of temporary contracts. (For that reason within Argentina, they are known as contratados.) Although in principle national parties could develop a cadre of potential bureaucrats, party fragmentation into provincial parties makes such cadre development ineffective in comparison with the practice of a parallel bureaucracy composed of individuals strongly aligned with the secretary or minister of the moment. Sometimes these loyalties correlate with the province of origin of the relevant political figure.18 The Leeway of the Executive

The previous sections have characterized a policymaking environment in which institutions (such as congress) designed to facilitate political debate, bargaining, and the intertemporal enforcement of agreements are weak; in which some key political players (the governors) care little about the quality of national policies; and in which complementary enforcement mechanisms such as a strong judiciary or a strong bureaucracy are also lacking. These factors are reinforced by, and reinforce—in a general equilibrium way—the capacity and tendency of the executive to act unilaterally. The sources behind the executive’s ability to effect unchecked moves and undo previous (say, legislative) agreements are various, inA visible recent example is provided by the top echelons of President Kirchner’s current administration, staffed by a proportionally large number of people from his sparsely populated province of origin, Santa Cruz. By the time of this writing, five out of eight state secretaries under the presidential office have been from the province of Santa Cruz, which contributes less than 1 percent of the country’s population. 18

105

106

Pablo T. Spiller

and

Mariano Tommasi

cluding the supreme court’s tendency to be politically aligned with the president, the lack of a strong and independent bureaucracy, and the “general equilibrium” result that congress has not built strong technical capacities. The budget process manifests this last feature. The inability of congress to monitor and control the budget has given the administration substantial budgetary discretion. Since the beginning of the twentieth century, congress has often failed to approve the budget sent by the executive on time, which in practice has meant that the administration functioned independently of congress. In years of high inflation, the executive often did not even bother to submit a budget. Even in the lowinflation period of the 1990s, although budgets began to be approved on time, congress did not exercise its ex post control. Indeed, the budget verification process (“Cuenta de Inversión”) has not traditionally been initiated promptly enough to be an operational instrument for congress to verify the fulfillment of the budget “contract” by the executive. Jones (2001, p. 161) indicates that almost all of the substantial budget activity happens in executive quarters and not in congress: “…relatively little modification of the budget proposed by the executive branch occurs at any time during the treatment of the budget bill in Congress.” Furthermore, Jones indicates that while ministries and other entities submit detailed disaggregated budget plans to the (executive) National Budget Office, the draft finally sent to congress contains expenditure only at a very macro level.19 Budget practices during the 1990s tended to overestimate revenues, and the (nonlegislative) follow-up adjustment mechanism gave substantial leeway to the secretary of the treasury to allocate scarce funds. Secretaries of the treasury have exercised the prerogative to allocate funds ex post according to a mix of their own whims and pressure from the various ministries, occasionally arbitrated by the president.20 A similar

See also Abuelafia and others (2005). Based on personal interviews by the authors with the then-secretary of the treasury (1999), Pablo Guidotti. See also Baldrich (2003), who in 2001 was secretary of the treasury, and Abuelafia and others (2005). 19 20

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

pattern has emerged during the Kirchner administration, with underestimation of revenues leading to discretionary executive allocation of the additional funds. The unilateral power of the executive has also been based on practices that have tended to grant the president more power in fact than he has on paper. A common practice has been that of congress explicitly delegating power over some legislation; that is in part a “general equilibrium” consequence of the lack of interest by legislators in policymaking. Another factor that has contributed in this direction is the history of democratic instability. The absence of a legislative power during the frequent military dictatorships of the twentieth century has tended to centralize the actions and expectations of multiple societal actors on the executive (Acuña 1995; De Riz 1986; Llanos 2002a). There are also explicit constitutional stipulations that have endowed the president with the capacity to “regulate” the laws from congress (“rulemaking” decrees in the course of implementing legislation); as well as constitutional vagaries, favorably interpreted by a weak supreme court, that granted the president the power to issue Decrees of Urgency and Necessity (DNUs) (Carey and Shugart 1998b; Molinelli, Palanza, and Sin 1999). DNUs are regulations dictated by the executive through which decisions are made that normally can only be taken through an act passed by congress. Before 1994, decree authority was not explicitly granted by the constitution, and legal experts disagreed over its constitutionality. These decrees are valid as long as they are dictated in response to situations of extreme gravity that endanger the continuity of the state or imply social breakdown. This ambiguous wording opens the way to all sorts of interpretations and therefore to presidential discretion, especially under subservient supreme courts. In the 130 years from the 1853 Constitution to the 1983 return to democracy, 25 DNUs have been counted; President Alfonsín signed 10 DNUs during his 1983–89 administration. It was President Menem who used and abused this legislative resource 166 times (or more, depending how one counts), a resource that was validated by a supreme court that he had packed (Mustapic 2002). That practice has continued in subsequent administrations.

107

108

Pablo T. Spiller

and

Mariano Tommasi

President Kirchner produced 193 DNUs in his first three years in office (2003–06).21 The ability of the executive to “purchase” policies (through fiscal federalism and other political channels) also depends on electoral results. Institutional actors such as congress and crucial political actors such as governors tend to follow one of two strategies: one is confrontational, making the life of the president as hard as possible; the other is to cooperate “blindly” with the executive in exchange for short-term favors. The choice between those two strategies tends to be guided by a combination of partisan and resource availability considerations. When the party in executive power does not coincide with the colors of governors (and congressional majorities), and when the (fiscal) resources in the hands of the executive are low and declining, a confrontational strategy (like the one used by Peronist governors and leaders during the demise of the De la Rúa government in 2001) is more likely. In other (more “normal”) times, exchanging votes for cash is the most common strategy. In either case, intertemporal cooperation necessary to build effective public policies is lacking. Calvo and Murillo (2005) provide a detailed analysis of the partisan channels for the exchanges of votes for fiscal resources. They show that Peronist presidents have an advantage at this game, given the better grip of their party on subnational politics and clientelistic exchange networks, as well as the connection of national congress to subnational power. Interestingly, this has led to a pattern that Calvo and Murillo have christened “a new iron law of Argentine politics.” A couple of decades ago, it was said that it was an iron law of Argentine politics that only Peronists could win the presidency in free and open elections. After the victories of Alfonsín in 1983 and De la Rúa in 1999, that law no longer holds, but might have been replaced by: “Others can win, but they cannot govern.” Counting DNUs is not a trivial exercise, since all types of presidential decrees receive a consecutive number, whether they are DNUs or whether they are other, more runof-the-mill presidential activities. For details on the different sources and figures see Braguinsky and Araujo (2006), Serrafero (2005), Ferreira Rubio and Goretti (1998) and references there. 21

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Argentina

Conclusion

This chapter began with the observation that Argentina’s policies have shifted dramatically over time, and with the claim that its inability to produce consistent public policies is the reason behind its economic misfortunes. We argued that the deficiencies of Argentine public policies are the outcome of a policymaking process in which key actors have little incentive to cooperate with one another over time, leading to myopic political and policy choices. Argentina has thus been unable to steer a consistent path in crucial areas such as economic, social, or international strategies. Argentina has a national policymaking environment dominated by executives who tend to have much leeway to pursue whatever policies they fancy, provided they can buy the support of quasi-feudal provincial governors with fiscal largesse. Such a depiction is somewhat surprising for a country that has a basic constitutional structure with separation of powers similar to that of the United States. In spite of a common basic constitutional structure, the workings of political institutions in Argentina could not be more different than in the United States. The Argentine congress is neither the arena where key policies are negotiated nor a very attractive place to develop a political career. The Argentine supreme court is not a respected institution with the power to act as a real check on executive power. In Argentina, there is no policy delegation to well-staffed independent professional agencies supervised by congress through the power of the purse. The Argentine president often seeks the support of provincial governors for his national policies, and provincial governors’ main objective is to increase the amount of federal funds they obtain to finance provincial public sector activities as well as their political machineries (both things often coincide). In such a scenario, interest groups such as business and unions also follow noncooperative short-term strategies, trying to maximize shortterm payoffs whenever they have access to state resources. This was clear during the “reform” experience of the 1990s, in which many reform measures amounted to a front-loading of payoffs to many key actors, such as local business groups favored by privatizations, key unions fa-

109

110

Pablo T. Spiller

and

Mariano Tommasi

vored by reforms in the health and pension system, and pension funds that charged high up-front commissions. So, socioeconomic actors also undertake shortsighted political strategies. Short-term maximization of political opportunities also shortens their horizons as economic actors. The short-term nature of most policy and institutional arrangements seems well understood by economic agents, who, for instance, did not contribute much to the “privatized” pension system. Some short-term credibility in monetary and exchange rate policy was bought through the rigid (and in the end, costly) convertibility system.22 In the 1990s Argentina undertook a rather surprising sharp turn toward market-oriented policies. According to the logic of policymaking described here, that turn was not the outcome of reasoned public debate in which most relevant political actors considered whether this policy shift was the most desirable course of action. It was a decision of the executive of the day, approved by congress through votes that were largely purchased through the federal fiscal system and related mechanisms.23 The implementation of these reforms carried the imprint of the noncooperative policymaking process described here. This was reflected in several peculiar characteristics of the policies of the 1990s, such as the rigidity of the convertibility regime, the inefficiencies of federal fiscal arrangements, the lack of enforcement of some policies, and the incoherence of privatization and regulatory policies across sectors. The overall experiment, including the convertibility regime, did not end well, at least as evaluated by the Argentine public at the time of this writing. The current Kirchner administration is fairly outspoken against several aspects of the reform process of the 1990s. It seems clear that unless some fundamental changes in the rules of the political game take place, Argentina will continue to experience low quality policies, independent of their political orientation. Etchemendy (2002), Murillo (1997, 2001), and Kay (2003) provide evidence on the short-term beneficiaries of the political exchanges behind market-oriented reforms (see also Acuña and others 2007). Galiani, Heymann, and Tommasi (2003) describe the political and economic dynamics behind the origin and maintenance of convertibility and its demise. 23 See Acuña and others (2007). 22

4 On the Road to Good Governance: Recovering from Economic and Political Shocks in Brazil Lee J. Alston, Marcus André Melo, Bernardo Mueller, and Carlos Pereira

Introduction

The Brazilian Constitution of 1988 defined the political institutions in Brazil and the powers of the political actors in the policymaking process following the end of the military government in 1985. As a result, 1988 is used as the point of departure for this analysis. In Brazil, strong presidential powers drive the policymaking process; though they are checked by the constitutionally defined powers of congress, the judiciary, state governors, and the Ministério Público (public prosecutors). Policymaking starts with an interaction between the president and members of congress, though it is always in the shadow of the other political actors. Significantly, the president’s electoral connection relies on a strong economy that is perceived as satisfying the goals of economic growth, economic opportunity, and the reduction of poverty. Members of congress generally care more about redistributing gains to their constituents, especially along geographic lines. Given the different preferences and relative powers of the legislative and executive branches, they can both benefit by exploiting the gains from trade. To achieve the goal of a strong economy, presidents focus first on fiscal and monetary stability (such as the Real plan), as well as on pension reform and tax reform.1 To achieve these ends, presidents Fernando Henrique Cardoso as finance minister put in place the Real plan, which led to price stability. The success of the Real plan was also instrumental in Cardoso winning his first presidential election. 1

112

Lee J. Alston, Marcus André Melo, Bernardo Mueller,

and

Carlos Pereira

have used pork as well as other benefits as the mediums of exchange with members of congress in return for their votes on critical pieces of legislation. A key element in this exchange is the allocation of selected ministerial positions and appointments in the bureaucracy. Given that a large proportion of the budget is “hard-wired” with policies such as pensions, health, and education, once the trades of pork-for-policy reform on monetary and fiscal policies have been consummated, the surplus will be spent on more ideological policies such as land reform and the environment. The residual policies have different electoral effects on the president and members of congress depending on the degree to which they achieve national goals (whereby the president can claim more credit) or geographic goals, such as poverty reduction in the Northeast (in which case the deputies and senators receive relatively more electoral benefits). This “game” can be viewed as sequential, with veto players as well as external shocks constraining the president and congress. Constraints from the other political actors and external shocks have budgetary implications, either positive or negative. This analysis views the policymaking “game” as one in which the president has an overriding incentive to keep the budget as a percentage of GDP within some target range, because greatly exceeding the budget to GDP ratio can have serious monetary and fiscal penalties imposed through international capital markets. This tendency to strive toward a target range of debt to GDP is called the budgetary equilibrium. The dynamics of the policymaking game yield policy outcomes that are classified as falling into four broad categories: “stable but adaptable,” pork, hard-wired, and residual. The second section of this chapter describes the outer features of public policies in more detail. The third section examines the political institutions and how they constrain the political actors. The passage of the 1988 Constitution is viewed as a “foundational” moment from which the president, congress, judiciary, governors, Ministério Público, and the Tribunais de Contas derive their powers.2 The third section also illustrates The Tribunais de Contas is analogous to the General Accounting Office in the United States. 2

the

Road

Good Governance

how the institutional players enter the game, constraining the executive and congress, and provides examples along with systematic evidence on the impact of the other players on policy outcomes. The fourth section uses the framework to analyze specific policy outcomes: stable monetary policy and the fiscal responsibility law, with stable but adaptable outcomes; land reform and other social policies, with volatile outcomes; and health and education policies, with rigid outcomes. The final section offers some concluding remarks about the overall policymaking process in Brazil since 1988. The Dependent Variable: Outer Features of Public Policies

The purpose of this study is not to analyze the details of any specific policy adopted in Brazil, but rather to explain the outer features of public policies: that is, the common characteristics that systematically permeate those policies. The analysis of these outer features follows the framework developed in the opening chapter. The political institutions determine the key players, the payoffs to the players, the arena in which they interact, and the frequency of their interaction. The characteristics of policies in Brazil vary markedly across specific policy issues. The policies have been separated into four categories according to those characteristics. The first category consists of policies that are stable and adaptable to shocks. These are basically macroeconomic policies, such as fiscal and monetary policies: that is, those with a direct impact on stabilization and economic growth. The second category involves policies used by the president to provide patronage to other political actors in exchange for support in approving his agenda of reforms: that is, geographically concentrated transfers, or “pork.” The third category includes policies that, having been hard-wired, cannot be easily changed and are consequently rigid and less susceptible to economic and political shocks. In Brazil, policies such as education and health that have national purpose and important second-round effects have been hard-wired. “Second-round effects” refers to the fact that these policies generate important positive externalities for society that are realized not in the short term but in future periods. Thus there is always a temptation

113

114

Lee J. Alston, Marcus André Melo, Bernardo Mueller,

and

Carlos Pereira

for shortsighted policymakers to postpone them in favor of policies in the first category, which is why at some point a consensus was reached to insulate those expenditures. The final category consists of residual policies, which include issues that are given priority only once the objectives of the policies in the first category have been secured. These are policies related to issues such as security, environment, poverty, and land reform. These policies tend to be volatile, oscillating according to political shocks, such as when a new president comes to office. In general it is expected that policies with a larger ideological component, such as land reform and poverty alleviation, will be in this group. However, infrastructure has also increasingly been treated as a residual policy. It is significant to note that Brazil is pushing strongly for expenditures on infrastructure to count as meeting fiscal targets. The analytical task is to show how the same set of political institutions leads to political transactions that result in public policies with the characteristics described in each of the four categories when mapping across specific policy issues with different features.3 The key features of Brazilian political institutions can be understood in terms of separation of power and separation of purpose (Haggard and McCubbins 2001). Although the details changed, the Constitution of 1988 maintained the notion of strong powers for the president inherited from the military dictatorship of 1964–85.4 The next section will present details of the political institu-

The point of departure of this analysis is the 1988 Constitution, which defined the current set of political institutions that determines players and their powers. Although this analysis technically covers the entire period since 1988, the fit is clearly stronger in the post-1994 period, which includes the two terms of President Fernando Henrique Cardoso, as well as the current term of President Luiz Inácio Lula da Silva. It is more difficult to see a pattern in the period from 1988 to 1994, since it covered the final year of the Sarney presidency (1989), the impeachment of the Collor presidency (1990–92), and the interim years of Itamar Franco (1992–93). During this time, the new political institutions were still in the process of being implemented and developed. Nevertheless, the model presented here applies to this period as well, since many of the subsequent changes were built on institutional changes that occurred during this time, which was also an important period for political players to define and learn the rules of the game. Therefore this study’s claim that presidents have incentives to pursue sensible macroeconomic policy is not invalidated by unorthodox behavior during this early period. 3

the

Road

Good Governance

tions that underlie those powers (including decree power, veto power, legislative rules, and the budgetary process) and evidence of their effects. This section simply states that strong presidential powers have generally allowed the president to initiate, pursue, and approve much of his policy agenda. Whereas such a scenario may seem perilous given Latin America’s history with strong presidents, Brazilian political institutions provide two sets of safeguards against abuse of those powers. The first is that the electoral connection for the president is such that he has incentives to pursue sensible macroeconomic policies, as he is seen by the electorate as being responsible for outcomes related to basic issues such as a strong economy, economic growth, and stabilization. Given the strong presidential powers, failure in these areas cannot be credibly blamed on other political actors such as congress or the judiciary. The second safeguard is that although the separation of powers is clearly biased toward the president, several other political actors with different motivations (separation of purpose) are able to check the president’s actions in different ways. Thus if an incompetent or ill-intentioned president were to come to power, strong presidentialism would not mean a blank check to pursue misguided policy. The next section will describe in detail the roles played in this balance of power by the other main players, particularly congress and the judiciary, and to a lesser degree, state governors, the Ministério Público, the bureaucracy, and regulatory agencies, showing that they can and often do constrain the president’s actions. The president uses his powers to pursue an agenda of stable and adaptable policies and reforms. Because the separation of purpose inherent in the political institutions has the president pursuing broad national public goods—as opposed to other actors with a say in producing legislation who have more narrow constituencies (particularly congress and to a lesser degree governors)—there is the potential for conflict. This resulted from the peaceful and voluntary transition from a military dictatorship to a civilian government. As a result, the Constitution of 1988 was written by congress, but with considerable input from President Sarney, who inherited strong presidential powers. It is logical that a strong military president would want to assure that the new constitution would not reduce those powers. In fact, one of his greatest concerns at the time was to be granted an additional year in office by the constitution. 4

115

116

Lee J. Alston, Marcus AndrĂŠ Melo, Bernardo Mueller,

and

Carlos Pereira

However, this conflict is diffused by legislative rules that result in trades of support for patronage between congress and the president (Alston and Mueller 2006). The president is able to use his powers to control the legislative agenda and to create a stable supporting coalition that enables policy reform. Any other coalition not coordinated by the president would be inherently unstable, as it would not have any enforcement mechanism to ensure compliance and prevent defections. In addition, the president possesses considerable discretion over patronage (such as jobs and individual budget amendments), which, together with the career incentives of members of congress (discussed in the third section), lead to the well-institutionalized trade of policy support for patronage. Although these exchanges are often seen as being less than legitimate by the press and much of society, they form the basis of executive-legislative relations in Brazil, and this study argues that they lead to high levels of governability that allow important reforms to get accomplished. Furthermore it can also be argued that this comes at relatively low cost to the executive because political institutions facilitate the trades (see discussion of individual budget amendments in the next section), and the patronage that is dispensed is a very small part of the budget (Pereira and Mueller 2002, 2003). The result is that congress tends to approve many of the policy reforms proposed by the executive, yet congress still holds checks on the powers of the president.5 It is this interaction that determines the qualities of policies in the first and second categories discussed above: stable macro policies on the one hand and pork on the other. Only when the divergence of preferences over specific policy issues between the president and members of congress is sufficiently high will there be no gains to trade. That is, the cost of the patronage necessary to approve those reforms is higher than the benefits to the president. This may lead to gridlock over that issue; presidential action to drop the issue or significantly water it down; or attempts by the executive to get around congress, for example

This interaction between the president and congress in Brazil is formally modeled in Alston and Mueller (2006). See Melo (2002) for the executiveâ&#x20AC;&#x2122;s recent success in amending the constitution. 5

the

Road

Good Governance

through decree power (discussed in the next section), which may then prompt other political actors such as the judiciary to intervene. The next sections argue and present evidence, however, that, except in a few highprofile cases such as pension reform during the Cardoso presidency and tax reform during President Lula’s term, the president has generally gotten what he wanted, with these other outcomes being exceptional. Depending on his success in achieving the top priority policy objectives, the president will decide which residual policies will be pursued, and how. Thus the residual policies are contingent on there being space in the legislative agenda, as well as on budgetary availability. These in turn are affected by both economic and political shocks. The fifth section provides one example of each shock: the effect of the 1999 devaluation of the real and the uncertainty resulting from the election of Luiz Inácio Lula da Silva (“Lula”) as president. In each of these cases it will be shown how the budgetary process was used to bring about a recovery by reducing the execution of the residual policies in the budget. This implies that many of these residual policies will be characterized by high volatility. Given this inherent tendency toward volatility of the residual policies, political actors may often choose to hard-wire some policies where it is considered that the volatility can be particularly damaging. This is especially the case with education and health policies that are crucial for social and economic well-being, since some politicians may nevertheless be tempted to withhold resources because the effects of these policies are not generally felt in the short term. Thus at some “constitutional moment,” politicians establish impediments to changing these policies by tying the hands of future political actors. This results in rigid policies, which are advantageous when this rigidity constrains opportunistic behavior but which comes at the cost of reducing the ability to adapt to unforeseen future contingencies. Political Institutions

This section provides an overview of Brazilian political institutions and how they affect the policymaking process. The section describes the key political actors, the payoffs for political cooperation, where and how fre-

117

118

Lee J. Alston, Marcus André Melo, Bernardo Mueller,

and

Carlos Pereira

quently the political actors undertake their exchanges, and the properties of the arenas in which exchanges take place. The discussion begins with an analysis of the way in which the 1988 Constitution established the current political institutions and how it has endogenously changed as an important mechanism for providing commitment for policy reform. Next the section describes the powers of the president and of congress and analyzes how their interaction generates policies and how the electoral and legislative rules influence the behavior of the president and members of congress. Finally the section concludes with an examination of other political actors that shape the policymaking process in many different ways, constraining what the president can do. The Constitution

The Constitution of 1988 established the rules of the current political game in Brazil. A Constituent Assembly convened in 1987 drafted the constitution. The Constituent Assembly was set up by conferring special powers on the ordinary legislature rather than by holding new elections for the purpose of writing a new constitution. The new constitution reflects a number of principles long advocated by the opposition: decentralization, transparency, participation, social control, and redistribution. These principles produced a significant transformation in the patterns of policymaking and implementation. In terms of fiscal and intergovernmental relations, the constitution devolved administrative autonomy to subnational governments and mandated a new redistribution of functional responsibilities. In addition, it mandated a new regime of tax assignments whereby the states and municipalities were given not only new tax powers but also managed to secure a larger share of federal tax revenues. The constitution created new funds for states and municípios by mandating automatic transfers of federal money. It also mandated the decentralization of public policy in a great number of sectors ranging from health to education to environmental policy. Furthermore, the constitution mandated multilevel participatory arrangements aimed at social control. They include an enhanced role for the Tribunais de Contas (Court of Accounts) and the Ministério Público

the

Road

Good Governance

(public prosecutors), as well as the decentralization of the judiciary branch. While the new Constituent Assembly was characterized by the strong desire to break away from a long period of authoritarian rule, a number of institutional innovations represented an element of continuity. Although the new constitution (by virtue of Article 2 of the provisional clauses) mandated a plebiscite on the regime and system of government that was set to take place in 1993, most of its features presupposed a strong presidency. The adoption of the constitution can be seen as a pivotal moment in time, with important path-dependent developments. Because it was formulated during a unique historical juncture, it incorporated a vast array of political, social, and corporatist demands that had been kept silent under centralized military rule. As a result, with 250 articles in the main text and an additional 75 provisional articles, the constitution is unusually long and covers many very specific nonconstitutional issues of policy. The decision to create such a wide-ranging and detailed constitution could be attributed to the lack of any sort of political trust and credibility at that time. Thus writing a constitutional article was a “safe” institutional method by which political players could undertake political transactions with some degree of certainty that their arrangements and agreements would be enforced. As a result, the constitution emerged with many policy issues hard-wired, meaning that changes would require constitutional amendments. Indeed much of the political capital of Presidents Cardoso and Lula was spent in deconstitutionalizing certain issues: that is, deleting articles from the constitution and subsequently (but not always) legislating issues through ordinary laws. The initial high level of constitutionalization of public policy produced great rigidity in public policy in general. This did not preclude the Cardoso or Lula da Silva administrations from passing their reform programs, however. The previous discussion helps explain why the Constitution of 1988 is the most amended constitution in the country’s history. Brazil’s first constitution, passed in 1824, lasted 65 years and was amended once. Brazil’s second constitution, which established the republican form of government, lasted from 1891 to 1930, and also was amended only once.

119

120

Lee J. Alston, Marcus André Melo, Bernardo Mueller,

and

Carlos Pereira

The Constitution of 1946 lasted 21 years and was amended 27 times. The military constitutions of 1967 and 1969 were amended 26 times in a period of 21 years. By contrast, the Constitution of 1988 was amended 37 times in 12 years. Between 1988 and February 2001, 2,424 constitutional amendments were presented to congress. Under President Lula da Silva’s first term, three new constitutional amendments were passed. The yearly average amendment rate for the Constitution of 1988 through 2003 reached 3.5. For the period from 1992 (when the first amendment was approved) to 2003, the average yearly rate of amendment was 4.4—an extremely high rate by any standard. These rates are all the more significant because constitutional change requires approval in two rounds of voting in each house by an absolute majority of three-fifths. Other procedural requirements include the following: the executive is not allowed to change the constitution by provisional decrees (medida provisória). Similarly, the executive cannot resort to special urgency procedures through which it could unilaterally require a vote on a bill ahead of any other legislative proposals. Moreover, the vote must proceed by roll call (thus increasing the political costs of approving unpopular proposals). The political transaction costs of securing legislative approval are therefore much higher for constitutional amendments than for ordinary legislation. Comparatively speaking, however, the requirements for approving constitutional amendments are not very strict; Brazil is in a cluster of countries whose constitutions are the most easily amended (Melo 1998). In addition to the procedural difficulties, it should be noted that constitutionalization and deconstitutionalization (inserting and deleting provisions from the constitution) are very distinct and asymmetrical political processes. For consitutionalization, collective action problems undermine the ability of the public to insert particularistic interests. For deconstitutionalization, the opposite holds; withdrawing benefits and rent-seeking privileges from the constitution requires overcoming the resistance of organized and sectoral interests. Reforms that deconstitutionalize issues initially lead to legislation regarding the issuing of a medida provisória. In this climate, a lack of trust and opportunistic behavior have precluded some welfare-enhancing deals from taking place. Two

the

Road

Good Governance

examples include tax reform initiatives involving the assignment of the power to collect value added taxes across local, state, and federal governments; and the elimination of taxes by states on exports. These potential policies were not implemented because no credible compensation mechanisms could be put in place by the federal government during a transition stage. The lack of trust arose because the executive holds great agenda powers and decree authority. Many members of congress saw the deletion of articles from the constitution as a mechanism by which the executive can unilaterally impose its preferences. In other words, some regard deconstitutionalization as the equivalent of giving a blank check to the executive. Many constitutional issues have a direct bearing on fiscal and financial stability, and therefore put the constitution in the center stage of the political game described earlier. These include the rules defining social security benefits, the provisions stipulating levels of pay and of hiring/ dismissing personnel, the stipulation of areas not open to foreign ownership, rules defining tax and fiscal matters of subnational governments, and central bank independence. Thus, we argue that the political game described herein is largely a “constitutional change game.” Granting constitutional status to certain policy areas is an integral part of hardwiring. As discussed below, this was the strategy pursued in the areas of education and health. In these cases, hard-wiring represented a strategy of precommitment on the part of the executive and the legislators. In this game, the supreme court plays the role of a veto power because of the institution of judicial review in the country. Two instruments can be used by players in the judicial review game: the ação direta de inconstitucionalidade (ADIN, a petition for nullifying a decision or legal norm because it is assumed to be unconstitutional) and the ação declaratória de constitucionalidade (a petition for the confirmation of constitutionality of a decision or legal norm). Both are to be decided by the supreme court (Supremo Tribunal Federal). The first type of petition can be initiated by the president, congressional parties, the attorney general, the speakers (mesas) of the senate, the chamber of deputies and of the state legislative assemblies, the governors, the bar association, trade unions, and professional bodies. The second type of petition can be initiated by the

121

122

Lee J. Alston, Marcus André Melo, Bernardo Mueller,

and

Carlos Pereira

president, the secretariats (mesas) of the senate or chamber of deputies, and the attorney general. The Executive, Congress, Parties, Committees, and Electoral Rules6

Since redemocratization, and especially after the new Constitution of 1988, all elected presidents have been able to build reasonably stable post-electoral majority coalitions within congress with a high level of governability by means of strong party discipline of the governing coalition (Figueiredo and Limongi 1999; Pereira and Mueller 2003). The only period without a stable majority coalition was from March 1990 to October 1992, under President Collor. Collor preferred to work through ad hoc coalitions and this undoubtedly added to the support for his impeachment. Although none of the elected presidents belonged to a party with a pre-electoral absolute majority of the seats, they have, nevertheless, been able to achieve congressional support by use of their extensive legislative and nonlegislative powers. Despite the presence of a decentralized electoral system and a fragmented party system, the optimal electoral strategy in the Brazilian legislature has not been concentrated in personal votes, but rather, the party vote in congress (Figueiredo and Limongi 1999; Pereira 2000; Nicolau 2000). At first glance, this assertion seems paradoxical, given the premise that legislators are subject to electoral incentives to behave individually. Indeed, Brazilian legislators vote according to their party leader’s indication in order to accumulate greater benefits in the congressional arena and thus to strengthen their electoral probability of political survival in the local sphere (Pereira and Renno 2003). This claim is also corroborated by Amorim Neto and Santos (2001b, p. 213), who argue, “Party New regulatory agencies and the activities of the Ministério Público are two increasingly important checks on the ability of the executive and congress to change policies quickly. Since 1997 Brazil has created 10 new regulatory agencies varying in their independence. The budgets of the Ministérios Públicos are hard-wired at both the state and federal levels, giving them considerable independence. Unfortunately, space constraints preclude a discussion of their roles. For more information, see the discussion in Alston and others (2007). 6

the

Road

Good Governance

discipline was above all a function of the Presidentâ&#x20AC;&#x2122;s legislative coalitionbuilding strategies based on dispersion of patronage to parties.â&#x20AC;?7 Scholars who analyze the Brazilian political system, especially its electoral rules and political parties, usually affirm that they provide significant obstacles for the executive to approve its agenda, thus creating tremendous governance problems (Mainwaring and Scully 1995a; Mainwaring and Shugart 1997a; Haggard 1995; Haggard and Kaufman 1992; Ames 1995a, 1995b, 2001). For these authors, the electoral rules offer strong incentives for candidates to develop direct links with their constituency groups rather than mediating such relations through political parties. Additionally, this institutional context generates incentives that lead to a personalized vote, as opposed to voting for political parties, and to a high saliency of constituency pressures in incumbentsâ&#x20AC;&#x2122; electoral calculus (Ames 1995a, 1995b; Samuels 2002). By contrast, a second group of authors has strongly questioned this predominant view. Rather than stressing the decentralizing effect of electoral rules, they emphasize the institutional rules and structures that centralize the legislative process itself and the powers held by the executive (Figueiredo and Limongi 1995, 1997, 1999; Meneguello 1998; Pereira and Mueller 2000). These authors attempt to explain how institutional variables internal to the decision-making process (the distribution of power inside congress) and the institutional legislative and nonlegislative powers held by the president (including decree and veto powers, the right to introduce new legislation, permission to request an urgency time limit to certain bills, discretionary power on the budget appropriation) work as key determinants for legislators to behave according to the preferences of party leaders. The Brazilian political system can be characterized neither as a purely decentralized nor as a purely concentrated system (Pereira and Mueller 2002, 2004). While some features such as electoral rules, a multiparty sysIn 12 consecutive elections (from 1950 to 1998) for the Brazilian chamber of deputies, the great majority of incumbents (70 percent, on average) have decided to run for reelection and almost 70 percent of them have been successful, more than most other countries in Latin America (Morgenstern 2002). This suggests that it is incorrect to ignore static ambition as the main goal among Brazilian legislators. 7

123

124

Lee J. Alston, Marcus AndrĂŠ Melo, Bernardo Mueller,

and

Carlos Pereira

tem, and federalism act toward decentralizing the political system, other features such as the internal rules of the decision-making process in congress, the constitutional powers of the president, and his or her capacity to selectively distribute political and financial resources (most of them locally allocated), act toward centralizing it. In fact, the electoral rules provide incentives for politicians to behave individually, while the internal rules of congress, the presidentâ&#x20AC;&#x2122;s power to legislate, and the centralization of benefits by the president render legislator behavior extremely dependent on loyalty to the party and on presidential preferences. It is claimed that even a political system with incentives for opposing behaviors, like the Brazilian one, provides equilibrium and stability. However, in this case it is a very dynamic equilibrium that can change from one issue to another and it depends on the capacity of the president and his party leaders to offer appropriate incentives (political and economic benefits) that can ensure the best electoral returns to individual legislators. This combination of institutional rules is key to understanding how it is possible for weak political parties in the electoral arena to coexist with strong political parties inside congress (Pereira and Mueller 2003). As a result, an underlying premise of our analysis is that there is no contradiction between party and individual behavior in the Brazilian political system at the same time. Legislators behave according to the preferences of party leaders within congress so as to have access to benefits that will increase their individual chance of surviving politically. Party leaders hold important institutional prerogatives: the ability to appoint and substitute members of committees at any time, to add or withdraw proposals in the legislative agenda, to decide if a bill will have urgency procedure, to indicate the position of the party regarding a bill on the floor, and, fundamentally, to negotiate with the executive the demands of the members of his party. In other words, party leaders are the bridges that link individual legislators with the preferences of the executive. This is why political parties are so strong within the legislative arena. It is not rational for legislators to act individually inside congress, just as it is not rational for the executive to negotiate or bargain for support with each member of his or her coalition on every bill. The role of intermediary between the executive and the individual

the

Road

Good Governance

legislators cements the fragile links between voters and representatives in the electoral sphere. Because the Brazilian political system works in the peculiar manner described above, one might observe a false contradiction between the “personal vote” and “party vote” approaches in the literature. But the two explanations are faces of the same coin. The most striking proactive power (which enables presidents to legislate and to establish a new status quo) in the Brazilian constitution is the ability of the president to legislate through provisional decrees (medidas provisórias). This institutional device allows the president to enact new legislation promptly and without congressional approval. Provisional decrees not only give the president the power to legislate, they also give him influence over the congressional agenda. If congress fails to act on a provisional decree within 30 days, it automatically goes to the top of the legislative agenda, displacing issues that congress may have been discussing. According to the constitution, a provisional decree should be used only in specific situations, although in practice the executive has made indiscriminate use of this device. Not only have a large number of provisional decrees been issued in past legislatures, but individual decrees have typically been reissued and amended at numerous times, since congress rarely challenges them. The supreme court tolerated this practice as long as presidents did not try to reintroduce any decree that congress had specifically rejected. In congress, serious disagreements over the extent of decree authority were not resolved until September 2001, when, in an accord with President Cardoso, congress amended Article 62 so as to limit presidents to a single reissue of a lapsed decree. The amendment also reduced constitutional ambiguity by specifying a list of issue-areas in which the executive may not resort to decrees. The partial rollback of presidential decree authority in late 2001 has altered the game of executive-legislative relations, and new patterns have yet to emerge (Pereira, Power, and Rennó 2005). The most common reactive power is the veto; it allows the president to defend the status quo by reacting to the legislature’s attempt to change it. The most common veto is the package veto, with which the president can reject the entire legislative bill sent by congress. Besides allowing the president to veto entire bills, the Brazilian constitution also allows partial

125

126

Lee J. Alston, Marcus André Melo, Bernardo Mueller,

and

Carlos Pereira

vetoes. The president may promulgate the articles of the bill with which he agrees, while vetoing the rest of the legislation and returning only the vetoed portions to congress for reconsideration. The 1988 Constitution makes it relatively easy for congress to override a presidential veto, given that an override requires only an absolute majority of the joint chambers. Nevertheless, the Brazilian congress has seldom used its veto power. This suggests that a majority of members of congress benefit from the status quo as compared to a counterfactual world of multiple parties facing a severe collective action problem in the legislative arena. In addition to provisional decree and veto power, the Brazilian constitution defines some policy areas where the executive has exclusive power to initiate legislation. Only the president can introduce bills concerning budgetary and public administration matters, as well as bills in an array of other important policy areas. In terms of budgetary law, although the congressional majority has the right to amend bills that were introduced by the president, it can do so only if those amendments are compatible with the multiyear budget plan elaborated by the executive, as well as with the law on budgetary guidelines. In addition, congress may not authorize expenditures that exceed the budgetary revenue. In practice, these rules enable the president to preserve the status quo on budgetary matters simply by not initiating a bill. The internal rules of the chamber of deputies give party leaders in the Steering Body (Mesa Diretora) and Board of Leaders (Colégio de Líderes) central roles in the legislative process and in the definition of the committee system. Roughly speaking, it is the prerogative of party leaders to appoint a committee’s members, as well as substitute them at any time (Article 10). There are no restrictions regarding how long a legislator can be a member of a committee. There may be some extent of self-selection to committee appointments, but there is evidence of significant interference by party leaders in the process of appointing and substituting committee members. Turnover of legislators from one committee to the next is typically extensive. Legislators change committees frequently, not only between years but also within years. Additionally the executive, through party leaders in congress, stacks certain committees with loyal members.

the

Road

Good Governance

Besides centralizing decision-making processes inside congress and allocating huge powers of legislating to the executive, the Brazilian political system also allows the president to control the distribution of political and financial resources. This provides colossal electoral consequences for those who have the chance to exploit them appropriately. In Brazil, the executive has exclusive power to initiate the annual budget. Although legislators have the right to propose individual amendments to the annual budget, the executive determines which amendments will be appropriated, making the budget contingent on the amount of available resources in the national treasury. The Brazilian president rewards those legislators who most vote for his interests by executing their individual amendments to the annual budget; at the same time, he punishes those who vote less frequently for his preferences.8 This is done by selectively executing their individual amendments (pork barrel policies). Legislators who are most successful in delivering pork barrel politics demonstrate a pattern of party behavior inside congress of consistently favoring the presidentâ&#x20AC;&#x2122;s preferences. To what extent has this legislative strategy been producing electoral returns? Pereira and Renno (2003) tested this question. They found that all other things held equal, the greater the amount of individual legislator amendments appropriated by the president, the higher will be the probability of a legislatorâ&#x20AC;&#x2122;s reelection. Pereira and Renno also found that the greater the number of individual amendments approved (but not appropriated) by the president, the lower is the probability that this legislator will be reelected. In other words, claiming credit is not enough to increase the chances of being reelected. The money has to be delivered. An additional result of the analysis indicates that there is no direct effect if a legislator votes in line with the wishes of the president; rather, the voting behavior matters for the indirect effect it has through pork. This leads to the inference that, in the electoral arena, the great majority of voters do not care about their representativeâ&#x20AC;&#x2122;s legislative behavior overall. Therefore, when legislators are deciding how they should vote on the floor, they are See Pereira (2000), Pereira and Mueller (2002, 2004), and Alston and Mueller (2006). 8

127

128

Lee J. Alston, Marcus André Melo, Bernardo Mueller,

and

Carlos Pereira

less inclined to take into consideration their constituency’s preference because it provides few benefits for their future political careers. Instead, their strategy is to access the benefits controlled by party leaders and by the executive. The Judiciary

Eleven judges serve on the supreme court. The president nominates judges for life terms, though with compulsory retirement at 70, and the senate confirms or denies nominations. The composition of the court has changed very slowly over time. Each president typically has the chance to appoint only a small number of judges, which makes it difficult to appoint the median voter in most issues, thus limiting the influence of the executive. The Constitution of 1988 further enhanced the independence of the judiciary by establishing that the judiciary determines its own annual budget, and the judicial courts appoint lower court judges. Both these rights removed potential instruments of control over the judiciary from other branches of government. The supreme court influences legislation both directly and indirectly: directly by ruling that legislation is unconstitutional, and indirectly by shaping what congress will consider passing. In other words, legislative and executive activities transpire “in the shadow of the courts.” If the supreme court were truly independent, then it should be possible to observe occasions whereby it directly contradicted the interests of the executive and congress. There have been a few high-profile cases in which the court ruled against the executive on issues that were of extreme importance to the executive. These are issues over which there can be no doubt of the executive’s preferences and will to prevail, so that if the judiciary were not truly independent, the executive would have used its power to change the court’s decision. The best example of this was the attempt by the Cardoso government to tax retired workers. The Cardoso administration envisioned this as an important component of the solution to the fiscal crisis of the government. The social security system in Brazil was seen as one of the main sources of the country’s large

the

Road

Good Governance

internal deficit. Taxing the transfers to retired workers proved highly controversial because it involved acquired rights and entitlements, and the executive was able to pass this measure through congress only with much effort (Alston and Mueller 2006). The supreme court, however, declared the measure unconstitutional. This decision enraged the administration and its supporters in congress. Initially, congress and the president threatened to deal with the supreme courtâ&#x20AC;&#x2122;s decision by changing the constitution. The reversal of the pension reforms also prompted a return to the episodic debate on the need for external control of the judiciary. This debate is revived every time a ruling by the supreme court or other parts of the judiciary gets in the way of governmental policy. Despite its threats to change the constitution, the Cardoso government abandoned the idea. The Lula government passed the same pension reform measure through congress in 2003. President Lula dealt with the potential challenge of the supreme court by making exceptions in the pension rules for the judicial branch and by counting on the support of the three members he had recently appointed.9 Given the independence of the supreme court, what can be said about its preferences and how can it be expected to act? Mueller (2001) analyzed the existence of commitment mechanisms for the government in the privatization and regulatory process in Brazil. His working assumption was that the court could be expected to act in a nonpolitical and unbiased manner in concession contracts, ruling closely to the letter of the contracts.10 The evidence is not yet in, but the current political debate over regulatory pricing is clearly taking place â&#x20AC;&#x153;in the shadow of the courts.â&#x20AC;?

Other important rulings by the supreme court include the mandate in 1997 by the court to increase the wages of federal civil servants to compensate them for losses due to previous stabilization programs, and the frequency with which the court prevented expropriations of land for land reform (see Alston, Libecap, and Mueller 2000). For a list and discussion of cases entailing conflicts between administrations and the court between 1988 and 1994, see Castro (1997). 10 This assumption does not mean that the Brazilian judiciary functions well in other aspects. Indeed there is evidence that the overall judicial system has a negative impact on the economy. See Pinheiro (1997) and Pinheiro and Cabral (1998). 9

129

130

Lee J. Alston, Marcus AndrĂŠ Melo, Bernardo Mueller,

and

Carlos Pereira

More generally, the main problem cited by companies, individuals, and judges themselves is the slowness of the judiciary (Pinheiro 2000).11 Courts at all levels, including the supreme court, are typically overloaded with cases, and decisions can take years. Companies often use this slowness strategically, taking actions that they know will be struck down in the distant future. For the government, this type of strategic behavior is even more appealing since the consequences of actions taken today may be left for future administrations. Governments have used the expected court delays successfully in tax legislation. If the courts rule against the government, it is future governments who will have to rebate taxes. In short, court delays act as a quick and cheap emergency means of government financing. Governors

Unlike the other actors discussed in this section, governors do not have an independent and constitutionally defined power that can directly counter executive preferences. Governors are not veto players, in the sense that their agreement is not necessary for approval of legislative proposals, nor do they hold powers to reverse legislative decisions. They can, however, have some indirect power over policy by their influence. The constitution vested a number of policy domains (such as public safety) to the states, which also enjoy certain autonomy in the area of taxes and administration. The policy preferences of governors and the executive may diverge in the political game described in the preceding section. Governors are not primarily concerned with fiscal stability at the national level and have a preference for higher federal public spending and geographically concentrated investments. Governors are also interested in more social transfers because these can be presented as state governmentâ&#x20AC;&#x2122;s spending and local programs. The preferences of governors and the executive clash over fiscal policy. It is necessary to distinguish two phases in the discussion of the ways in which governors, and federalism at large, have constrained the execu11

Another well-cited problem is the lack of access to the judiciary by the poor.

the

Road

Good Governance

tive. The first phase was transitional, when the rules of the political game were not yet fully institutionalized. In this period, governors derived their power by virtue of the role they played in the democratic transition. The second phase dates from the Cardoso administration, when the new constitutional rules of the game were in place. Governors have derived their powers from two sources. First, as previously mentioned, they enjoyed great political power in the 1980s because of the role they played in the democratic transition. In addition, while there is much dispute in the political science literature, governors tend to have some, albeit declining, influence over the behavior of federal deputies and senators in congress. The degree of influence they hold varies across areas, with governors playing a crucial role in issues with important state-wide effects, such as tax and regional infrastructure. Governors can also play an important role in the electoral career of legislators at the state level. Samuels (2003), for example, claims that congressional candidates tend to coordinate their campaigns around gubernatorial candidates and not presidential candidates. That is what he calls the “gubernatorial coat-tails effect,” in which the race for governor shapes the race for federal deputy because Brazilian politicians do not obtain much of the electoral resources they need from national parties or presidential candidates, but from statelevel connections. However, the broker’s role attributed to governors and their control over their state’s legislative delegations have been greatly exaggerated by the media and by the literature, especially during the Cardoso administration. It is true that governors can constrain the executive in indirect ways, but they do not wield veto power in any federal arena. There is no evidence that state loyalties on the part of legislators undermine party lines or create trouble for the executive (Cheibub, Figueiredo, and Limongi 2002). The indirect ways in which governors influence policy formation range from lobbying activities on specific issues affecting states, such as tax policy, to control over specific appointments in the federal bureaucracy. Governors can also resist the implementation of federal policy, such as when a number of state governors refused to privatize state-owned energy firms. These episodes are unusual and have become

131

132

Lee J. Alston, Marcus André Melo, Bernardo Mueller,

and

Carlos Pereira

even more so because states have become less and less autonomous and more dependent on the federal government. Another declining institutional source of a state’s power has to do with the state’s prerogative to own banks and public enterprises. The state banks were created in the 1960s as part of the developmental strategies pursued by the military government. With the democratization of the country in the 1980s, the governors became more autonomous from the federal government and therefore were able to use the state banks for pork. This included granting subsidized loans to the private sector and more importantly, financing state government projects that are fiscally unsound. The state treasuries also issued bonds that were purchased by the banks. Thus during the Sarney, Collor, and Itamar administrations, the states operated under a soft budget constraint because of their ability to undermine the supervision of the central bank. In 1994, before the establishment of the privatization program (PROES), there were 35 state banks. Currently, there are four small state banks. In addition to the state banks, governors control a vast network of sources of pork, ranging from public sector jobs to infrastructure programs and state-owned enterprises. In state-owned enterprises, a pattern similar to what occurred in the banking sector can be observed. Most public utilities companies in energy were privatized, leaving governors without an important instrument of power. The resources controlled by the states are important assets that are instrumental to winning elections, and they are coveted by the federal government. In the large states, the administrative apparati are large machines that can be even larger than the federal government machine itself, as in São Paulo. Because of the fiscal problems facing the states following monetary stabilization, the federal government was able to impose the privatization of the banks and public enterprises, thereby dramatically undercutting the power of governors. With inflation under control, the state banks lost their principal source of revenue (the floating of financial assets), and a surge in interest rates caused a rapid deterioration of the states’ fiscal situation. State debt reached a peak in 1997 (three years after the Plano Real was phased in) and represented a significant part of the GDP. The Plano Real was therefore an exogenous shock that undermined the

the

Road

Good Governance

ability of the states to resist the preferences of the executive. The federal government implemented a program aimed at renegotiating the states’ debts. It included debt swapping under favorable conditions, though it was linked to a number of conditionalities.12 Before 1994, a number of incentives such as federal bailouts encouraged states to behave fiscally irresponsibly and indulge in opportunistic behavior. In sum, while it is clear that federalism matters and that governors play an important role, throughout most of the last decade the executive has been able to have its agenda implemented by recentralizing the political game. This includes passing legislation that adversely affected the state governors and initiating measures that have led to a political recentralization of the country (Melo 2002).13 The Bureaucracy

This chapter views the Brazilian bureaucracy as an important institutional player and institutional constraint in the political game. The bureaucracy is an institutional actor that constrains the executive but at the same time plays an integral part in the management of the government coalition. The bureaucracy has been undergoing significant change in response to the changing economic and political environment. In particular, the role of the bureaucracy within regulatory agencies is increasingly important. Overall, the changes in bureaucratic structures following the enactment of the Constitution of 1988 have led to policy outcomes that are stable and adaptable. Brazil has been discussed in the literature on comparative bureaucracy as a fairly successful case. During the so-called Estado Novo (1937–45), President Vargas implemented a significant administrative reform. It set up the Departamento Administrativo do Serviço Público (DASP, or Adminis-

This was accomplished by the Programa de Recuperação Fiscal e Financeira (RFF) and through Law 9496 in 1997. A stock of debts corresponding to 11 percent of the Brazilian GDP was renegotiated (Mora 2002). 13 For a contrasting view that argues that resistance from the states and governors undermines much of Cardoso’s reform efforts, see Samuels (2003). 12

133

134

Lee J. Alston, Marcus AndrĂŠ Melo, Bernardo Mueller,

and

Carlos Pereira

trative Department for the Civil Service) in 1938 as the key administrative agency responsible for the competitive selection of federal personnel and for the rationalization of administrative practices and procedures. The DASP reforms led to the formation of a hybrid, two-level administrative structure in Brazil. The first level consisted of the core developmental bureaucracies in agencies in state-owned enterprises, state-owned banks, and in planning, taxation, and budgeting. The second level consisted of the administrative structures of the line ministries, particularly in the social sectors. The first level was insulated from competitive politics. The second level was part and parcel of patronage games and highly clientelistic arrangements (Geddes 1995; Nunes 1997). Examples of the first type of insulated bureaucracies in the 1950s include the National Bank for Economic Development (BNDES), the Bank for the Brazilian Northeast (BNB), the Bank of Brazil, the Brazilian Institute for Geography and Statistics (IBGE), the Northeast Development Agency (SUDENE), and the Ministry of Foreign Relations (Itamaraty). Despite the insulation of some bureaucracies, the president still retains a large degree of power to make appointments, especially at the cabinet level. As noted earlier, the president uses both pork and patronage as ways of solidifying the support of his coalition. In order to obtain his preferences for fiscal stability, the president faces the dilemma of delegating bureaucratic discretion to coalition party members while reducing the associated agency losses. The institutional rules governing the bureaucracy have enabled presidents to successfully play this game. The president can resort to 18,000 political appointments (known as DAS posts); a considerable number of these are low-rank posts. The key high-rank posts (approximately 3,000) are filled by the DAS 4, 5, and 6 appointments, representing less than 2 percent of federal public employees. Presidents have recruited personnel for these positions from within the civil service, from nontenured but highly qualified professionals currently holding positions in the bureaucracy, from public universities, and from the private sector. The president delegates less in the areas of taxation, budgeting, and planning in the ministries of finance and planning. Top-rank bureaucrats in these ministries are typically appointed from a pool of career

the

Road

Good Governance

civil servants in the central bank, the Internal Revenue Service, and the Itamaraty, among other institutions, and less frequently from outside government. Unlike countries such as the United States or France, appointments are made across career lines (Loureiro and Abrucio 1999). The president can then combine distinct criteria while making these appointments. These include personal loyalty and technical expertise. The latter is assured in these careers by extremely competitive entrance examinations and subsequent training in a number of civil service schools, including the Escola de Administração Fazendária and Escola Nacional de Administração Pública. The rest of the ministries have a less endogenous source of recruitment, and ministerial posts are assigned on a partisan basis. Presidents have usually kept the prerogative of appointing the ministries’ secretarygeneral—second in line to the minister, and in charge of managing the ministries’ positions—as a mechanism for reducing agency losses. The ministry of finance, however, is a key institution in this regard. By controlling budget execution and the cash flow of government expenditures, presidential control over the ministry is crucial (Loureiro and Abrucio 1999). Critical factors that explain the ability of bureaucratic executives to ensure a reasonable level of technical expertise in the Brazilian federal bureaucracy include the following: the widespread use of competitive entrance examinations in the areas of tax administration, budgeting, control, economic planning, accounting, central banking, social security, and legal positions within the executive; and favorable employment conditions in the public sector. These include tenure and reasonably competitive salaries. In the 1990s public employees’ real salaries eroded, but they were raised significantly during the Cardoso administration. The government also managed to change the civil service rules enshrined in the Constitution of 1988. The constitution introduced important changes in the Brazilian administrative state. The extension of tenure to all state employees (formerly called CLT workers) through Regime Jurídico Único (the Unified Legal System, or RJU) created a rigid system of personnel that exacerbated state inefficiency. The RJU prohib-

135

136

Lee J. Alston, Marcus AndrĂŠ Melo, Bernardo Mueller,

and

Carlos Pereira

ited different pay levels for distinct performance levels by state employees. It established the principle of equal pay for categories of functions at the municipal, state, and federal government levels. It also granted tenure and a secured 100 percent replacement rate for civil servant pensions (in several cases, up to 130 percent). The 1988 Constitution also created or strengthened bureaucratic and time-consuming mechanisms for competitive bidding and personnel recruitment, thereby creating an incentive structure that encouraged inefficiency. The administrative reform package was approved in 1998 and included revamping the RJU, setting up the legal foundations for social organizations and executive agenciesâ&#x20AC;&#x201D;institutions with managerial autonomy and with social control mechanisms, establishing performance contracts within the public sector, and making tenure more flexible. A number of measures also boosted the attractiveness of public employment: the ratio between initial pay and top grade pay within specific civil service career tracks was expanded significantly. The reform aimed at the low and middle ranks of the bureaucracy, where the pay is high, performance is poor, and the fiscal costs are very significant because of the sheer numbers of employees. Another important positive development under the Cardoso administration was the extension of professionalization outside the core of the economic, planning, finance, and infrastructure ministries. The ministries for the social sectors, particularly the ministry of health and education, also underwent important changes. For the first time, the ministers, secretary-general, and key managers were economists and were much more qualified than their predecessors. Policy Outcomes

Rather than generating consistent policy outcomes, the political institutions in Brazil generate policies in four broad but distinct categories: stable but adaptable policies; pork, or geographically distinct projects; volatile and unstable policies; and rigid and hard-wired policies. Consistent with this studyâ&#x20AC;&#x2122;s claims that the government has incentives and instruments to pursue sound fiscal and monetary policy, the first subsection below de-

the

Road

Good Governance

scribes those policies and the recent Fiscal Responsibility Law. Pork is also examined in this context, because it is often given by the president in exchange for other actorsâ&#x20AC;&#x2122; support on certain stable but adaptable policies. Then the discussion turns to the volatile and unstable policies, including social and poverty policies, as well as land reform. Finally, education policy and health policy are analyzed. These are examples of policies that are hard-wired, given the long lag that exists between spending and outcomes in these sectors, which may tempt policymakers to divert spending to other, less fundamental areas that have more immediate dividends. Stable and Adaptable Policies, Along with Pork

Stable but adaptable means that should outside conditions be stable, policy in these areas would be automatic. But exogenous events occur sporadically and unpredictably, so the government adapts to the events so as to minimize the damage to fiscal and monetary stability. The policies that best fit into this category relate to economic growth, inflation, and unemployment. To achieve policies such as pension reform, the president exchanges pork for policy support with members of his coalition. The following subsections first discuss why we consider macroeconomic policy to be generally stable and adaptable, even though economic performance has not been stellar (although recently growth has increased). The discussion then turns to the Fiscal Responsibility Law, which reined in state government debt. Macroeconomic Policy and Economic Performance. The framework presented in this chapter states that political institutions give the executive incentives to be concerned primarily with macroeconomic policy. In addition, strong presidential powers give the executive the means to pursue those policies which, when combined with the incentives and checks by other political actors, result in stable and adaptable policy. Meanwhile other policies, described in the next section, remain contingent on the success of macroeconomic policies to be executed. The framework presents a rather positive picture of the Brazilian policymaking process, especially when compared to most other Latin

137

138

Lee J. Alston, Marcus AndrĂŠ Melo, Bernardo Mueller,

and

Carlos Pereira

American countries. Although there are clearly several problems with this policymaking process, we maintain that overall the process provides a reasonably good means for intertemporal political transaction to be realized. The result is a system of checks and balances where a strong president achieves high levels of governability, positively constrained by several other political actors. These claims, however, may seem to many observers to clash with actual facts. An examination of real GDP growth rates shows economic growth has been less than spectacular: 0.7 percent on average for Brazil from 1990 to 1999, less than the average of 1.4 percent for Latin America as a whole.14 In addition, public sector debt, which is a key variable for gauging the sustainability of public policy, presents a trend that is a serious cause for concern: since 1999 the ratio of government debt to GDP has been above 50 percent. However, we argue that legitimate political change does not translate instantaneously into economic growth: policy changes create disruptions, and it takes time for the positive growth results to materialize. The growth record for the twenty-first century, especially in the past few years, has been a good indication that past policies are having a positive impact. This chapter does not claim that economic outcomes have been as positive as can be desired, but rather that the underlying policymaking process has had some very positive characteristics. Clearly a functional policymaking process should lead on average to good economic outcomesâ&#x20AC;&#x201D;although the fundamental link between these two is not immediate, since history (path-dependence) and other intervening factors may delay the effect of new political institutions on economic outcomes. The policymaking process portrayed here has evolved gradually over time and is still evolving. Many of the positive incentives tied to that policymaking process, which enable political transactions to be realized, have been functioning for only a relatively short period of time. The ma-

In March 2007 the Brazilian census bureau released revised figures for past GDP growth, incorporating improved methodologies in the calculation. The new numbers showed markedly better outcomes. GDP growth from 2000 to 2005 was revised from 4.4 to 4.3 percent (2000), 1.3 to 1.3 percent (2001), 1.9 to 2.7 percent (2002), 0.5 to 1.1 percent (2003), 4.9 to 5.7 percent (2004), and 2.3 to 2.9 percent (2005). 14

the

Road

Good Governance

jor process of institutional change that has led to this system started only in 1985 with redemocratization, and especially with the new Constitution of 1988. However, these changes did not come into effect immediately. Even after the promulgation of the constitution, there was still a long period in which complementary laws were being created and voted on in congress. More importantly a process of gradual changes has occurred, as discussed above. Furthermore, in the first years after the adoption of the constitution, the political process underwent a convoluted period because of shocks not directly influenced by the new institutions. Most of the important institutional changes began only in 1995, which is why most of our analysis focuses on the Cardoso and Lula administrations. Since 1995, many of the formal and informal rules that currently permeate executive-legislative relations were consolidated and became routine. In addition, several important reforms began to be implemented, representing not only important policy outcomes, but also altering the nature of the policymaking process by changing the political actors’ incentives and constraints: that is, shaping the political institutions themselves. The following changes have been among the most important: administrative reform, which changed the rules governing civil servants; privatization and the creation of regulatory agencies in several sectors; passage of the Fiscal Responsibility Law, which constrains political actors, especially governors (see discussion below); the evolution of the role of the Ministério Público, which has become an important veto player; and deconstitutionalizing many policies originally embedded in the constitution, such as pension reform. Yet to come is reform of the judiciary, already initiated by the Lula government. Fiscal Responsibility Law (FRL). Congress and the president enacted the FRL in 2000. The FRL represented the apex of a relatively successful set of measures to control the state governments’ indebtedness. The FRL illustrates the kinds of policy outcomes that reflect the executive’s ability to implement its policy preferences in the political game, discussed in the previous sections. Furthermore, it reflects a learning process arising from a repeated game between the federal government and the states.

139

140

Lee J. Alston, Marcus André Melo, Bernardo Mueller,

and

Carlos Pereira

As Braun and Tommasi (2004) point out, fiscal rules to be enforced require self-enforcement by the players (states) or an external enforcer with the power to ensure compliance. We argue that the Brazilian case approximates the second case. As discussed in this chapter, the current depiction of the Brazilian political system as a federal structure in which governors wield vast powers is inaccurate. In fact, the circumstances that originally produced strong powers at the state level were extraordinary: a Constituent Assembly in which the executive played a minor role; the political conjuncture of the transition to democratic rule, in which fiscal decentralization and increased social spending were important banners; and the specific sequence through which the political transition (democratic elections) occurred first at the state level (1982) and subsequently at the national level, converting the governors into key political figures in the transition. However, unlike pre-1994 Argentina, the political survival of the president or of the senators does not depend on subnational institutions such as the electoral college, in provincial assemblies in which governors play a key role.15 Because there was no constitutional basis for the power of governors, the Real plan represented a shock that restored the dominant power of the president. Among other things, it laid bare the states’ fiscal imbalances, made it impossible for the states to resort to floating and other financial mechanisms to finance their fiscal deficits, and caused a further deterioration of the deficits because of the sharp increase in interest rates. The executive was able to impose its fiscal preferences because it could offer advances from the federal development bank, BNDES, in exchange for fiscal reforms, including privatization of state banks and utilitiesz it had agenda powers and other legislative prerogatives to implement its agenda; and it was helped by the approval of the reelection amendment, which strengthened not only the president vis-à-vis the governors, but also helped extend the governors’ time horizons (19 governors ran for reelection), thus introducing some Current views on fiscal rules fail to recognize these crucial differences and instead categorize Brazil and Argentina as examples of the same perverse fiscal federal game (Melo 2004). These contributions fail to recognize the great preponderance of the executive in the fiscal game. For analyses of the Brazilian case along these lines, see Rodden (2003), Braun and Tommasi (2004), and Webb (2004). 15

the

Road

Good Governance

element of self-enforcement in the fiscal game.16 In addition, because of the devastating impact of hyperinflation in the mid-1990s, the presidentâ&#x20AC;&#x2122;s policies were viewed favorably by a great majority of the public, which became strongly inflation-averse. The sustainability of the current fiscal situation is therefore not dependent on the statesâ&#x20AC;&#x2122; cooperation. Although the FRL could be reversed, there is some rigidity in it, since a three-fifths majority in two rounds of voting in the two chambers is required for a change in the law. The FRL specifies in great detail the fiscal rules governing public sector indebtedness, credit operations, and public accountâ&#x20AC;&#x2122;s reporting. The law prohibits the federal government from financing subnational governments, therefore eliminating the possibility of bailouts as well as any changes in the financial clauses of the existing debt-restructuring agreement. The FRL imposes debt ceilings for each level of government. The executive branch proposes the ceilings and the senate must approve. The law stipulates that in the context of economic instability or drastic changes in monetary or exchange rate policy, the federal government can submit to the senate a proposal for changing these limits. Any excesses to the limits are to be eliminated within one year, otherwise new financing and voluntary transfers from the central government are prohibited. Other sanctions include withholding federal transfers by the federal government, denial of credit guarantees, and banning of new debt (Nascimento and Debus n.d.). In addition, the financing arrangements are transparent to all parties, including the public, and the FRL contains a golden rule provision for capital spending, that is, annual credit disbursement cannot exceed capital spending. It would be misleading to conclude that the impressive fiscal costs to the central government meant that the initiatives were in the interest of subnational governments. Furthermore, the fact that most of the fiscal adjustment was generated by raising tax revenue rather than by significant cuts in expenditures does not mean that there has not been a radical change in the intergovernmental balance of power. The states had to Without the reelection amendment, incumbent governors would have an incentive to exacerbate the common pool problem by leaving the fiscal problem to future governors. 16

141

142

Lee J. Alston, Marcus AndrĂŠ Melo, Bernardo Mueller,

and

Carlos Pereira

privatize or close their banks, embark on a program of fiscal modernization, reduce the relative importance of payroll (for which governors were required at least to refrain from hiring more personnel), sell enterprises, as well as to adapt their pension regimes to the federal rules (in addition to being prohibited from creating new pension institutions or legislating in this area). In sum, the states lost significant degrees of autonomy. As Mora (2002) argues, the states have become more and more dependent on the federal government because voluntary (that is, discretionary) transfers from the federal government have become essential for their fiscal survival. Many observers have praised the FRL because it laid the foundation for a new formidable system of rules for fiscal management. The IMF (2001, p. 1) described the new fiscal institutions as follows: In the last few years Brazil has achieved a high degree of fiscal transparency, together with major improvements in the management of its public finances. This was done against the background of an international and domestic macroeconomic environment that has posed substantial challenges to the countryâ&#x20AC;&#x2122;s economic policymakers. The cornerstone of these achievements has been the enactment in May 2000 of the Fiscal Responsibility Law, which sets out for all levels of government fiscal rules designed to ensure medium-term fiscal sustainability, and strict transparency requirements to underpin the effectiveness and credibility of such rules. Another pillar of the improved fiscal management has been the medium-term expenditure framework aimed at better aligning the allocation of budgetary resources over time to the governmentâ&#x20AC;&#x2122;s priorities and regional development strategy. Also instrumental in promoting sustained fiscal adjustment of sub-national governments has been the firm enforcement by the federal government of the debt restructuring agreements with most states and many municipalities. Volatile and Unstable Policy

We categorize certain policies as being volatile and unstable. Policies are unstable because some have a strong ideological component and as

the

Road

Good Governance

such will oscillate with changes—in the executive branch, in particular; and some are residual, in that the appropriations are determined so as to meet a budgetary target that does not upset the overarching goal of stable monetary and fiscal policy. The mechanism by which receipts and expenditures are balanced by the government throughout the budgetary year so as to achieve the target primary surplus is known as contingenciamento. At the beginning of each fiscal year, the government passes a decree impounding part of the discretionary expenditures in the budget: that is, those pending expenditures that are not hard-wired. As the year proceeds, these resources can be “unimpounded” if tax receipts are greater than expected and if hard-wired expenditures have not been greater than expected. The size of the cuts are set for each ministry and it is then up to each minister—together with the executive, but without consulting congress— to determine which programs and projects will be hit. Because these are necessarily in the “investment” part of the budget, which is the only part that is not hard-wired, these projects typically affect the policies that we label volatile residual policies. They become volatile because they undergo a process where they will be executed more fully in good fiscal years than in bad fiscal years. The next two subsections will discuss two sets of policies that fall into this category: land reform and poverty alleviation. Land Reform Policies. Land reform is the prototypical social policy. It is perhaps the most ideologically charged policy issue in all of Latin America. In addition, it is the kind of issue where the economic benefits of well-conducted policy materialize only in the long run. That is, land reform has those characteristics that the framework presented here predicts will lead to volatile policy, with changes in emphasis, design, and implementation coming about with each change in government. In Brazil, this is clearly a predominant characteristic of land reform policies. Since the 1960s every government has had a specific land reform program, although more than 40 years of effort have not managed to budge the indices of land ownership concentration. Each new administration has created a new land reform program with a different name and has set ambitious targets of how many families

143

144

Lee J. Alston, Marcus André Melo, Bernardo Mueller,

and

Carlos Pereira

it planned to settle. The executive includes resources in his budget proposal to congress and congress typically adds amendments to increase expenditures. But at the beginning of each fiscal year, the executive issues a decree that impounds (contingenciamento) part of the resources from several areas, so as to assure that the government’s expenditures and receipts are compatible with the primary fiscal surplus target, as described above. For example, the percentages of the budgets for land reform that Presidents Cardoso and Lula impounded from 2001 to 2004 were 11, 25, 36, and 6 percent. In short, the budgets for land reform demonstrate considerable volatility over time and within a term of a given president.17 As predicted by this chapter’s model, the Lula government cut expenditures on land reform, a residual ideological issue, despite its ideological preferences for greater spending on land reform. The rationale for the cut was to meet the fiscal imperatives of a particularly troublesome year. Poverty Alleviation. Brazilian antipoverty programs exemplify the kinds of policy areas that have exhibited highly unstable patterns in the last decade. According to the political game discussed earlier, they are residual policies. There is ample evidence of instability in the period from 1988 until the end of the Cardoso government. The evidence during Lula’s administration is more mixed and hinges on the interpretation of the high visibility received by the flagship program Fome Zero (zero hunger). The Brazilian constitution stipulates a clear role for the state in terms of poverty alleviation. According to Article 23 of the constitution, poverty alleviation is the joint mandate of the federal government, the states, and the municipal governments. In the 1980s there were two federal institutions aimed at reducing poverty: the Legião Brasileira de Assistência (LBA) and the Centro Brasileiro para a Infância e Adolescência (CEBIA). The Collor administration (1990–92) transferred many of the programs to a newly created Ministério do Bem Estar Social and also closed a large number of programs. In the area of food and nutrition, the administraBudgetary data, including that on impoundments, comes from several technical notes of the Consultoria de Orçamento e Fiscalização Financeira of the house of representatives, http://www.camara.gov.br/internet/orcament/Principal/exibe.asp?idePai=16&ca deia=0@. 17

the

Road

Good Governance

tion terminated the eight subprograms that had existed in the 1980s from which children and expectant mothers had benefited (Resende 2000). At the same time, President Collor phased out the Programa de Alimentação do Trabalhador (PAT), a food program for low-income workers. Upon taking office, President Itamar Franco announced poverty alleviation as one of his top priorities, dramatically changing course from the Collor administration. IPEA, the planning ministry’s economic think tank, prepared a “map of hunger.” The announcement of the map and the publication of figures pointing to the existence of 32 million people living in extreme poverty led Franco to declare the country to be in a “state of social calamity.” The next step was the announcement of a number of emergency measures and the setting up, in 1993, of the Conselho Nacional de Segurança Alimentar (CONSEA, the National Council for Food Security), consisting of 8 ministers and 21 representatives from civil society. Under President Cardoso, Comunidade Solidária became the main antipoverty initiative. The programs consisted of public and private partnerships and included volunteer groups. Overall, Comunidade Solidária functioned as a coordination mechanism designed to take place on two levels: it would encourage and facilitate the participation of civil society institutions in the formulation and implementation of social assistance programs, and it would identify current social spending programs that had a higher impact on poverty and channel resources to those programs. The programs identified received the “priority seal,” which protected them from expenditure cuts. The Lula administration initiated a reversal of Cardoso’s antipoverty programs, at least on paper. The Lula administration phased out Comunidade Solidária and launched a new antipoverty program, under the auspices of the Ministério Extraordinário para a Segurança Alimentar e Combate a Fome (MESA, an ad hoc ministry for food security). The Fome Zero program became the centerpiece of the new ministry. However, President Lula reduced the resources earmarked for the program because of his fiscal targets.18 At the same time, the government decided The budget law for 2003 contained R$1.8 billion for the Fome Zero, whereas the budget estimate for its full implementation was R$5.0 billion. 18

145

146

Lee J. Alston, Marcus André Melo, Bernardo Mueller,

and

Carlos Pereira

to introduce a single card, the cartão família, which was to be used by families to receive the food benefit plus three other current conditional income transfer benefits. To receive the benefits, families must engage in a number of activities, including vaccination and school attendance. In practice, the government discontinued its initiatives and promoted the merger of cash transfer programs created under President Cardoso: the school attendance program (Bolsa Escola), food aid (Bolsa Alimentação), and the gas benefit. By creating the card, the Lula government could present the achievements of Fome Zero—that is, the number of families that received the card—in the same package with the figures for the much larger existing programs. The same applies to the funds allocated to the family card. The net result was that the political visibility of the modest resources allocated to the food benefit was reduced. Before Lula’s administration, during President Cardoso’s second term, an attempt was made to hard-wire funds for poverty alleviation. While acknowledging the gravity of the poverty situation, the Cardoso administration opposed the idea of a fund because it would imply “budget rigidity.” The prima facie attempt to introduce rigidity in the budget for the purposes of poverty alleviation can be understood as part of the logrolling between the executive and legislative branches. The executive maintained fiscal stability by increasing taxes at the subnational level. In return, congress received some poverty alleviation programs sheltered from discretionary executive budget cuts. This is consistent with the political game described in this chapter. The executive prefers hard-wiring subnational spending so as to allow for fiscal discretion at the national level. Rigid or Hard-Wired Policy

This section provides examples of hard-wired policies. These are policies specified in the constitution over which the executive has no discretion. In 2003, approximately 94 percent of the expenditures in the budget were “rigid,” that is, they could not be changed. The largest portion of these expenditures included types of expenditures whose shares grew

the

Road

Good Governance

over time and whose amounts were most probably not foreseen at the time of hard-wiring. The best example is pensions, which now account for 33 percent of the budget. Other types of hard-wired expenditures occurred intentionally; they are the outcomes of the deals between the legislature and the executive. They include transfers to states and municipalities (18 percent), the Unified Health System or SUS (7 percent), miscellaneous subsidies, including subsidies for small farmers and small enterprises, rural development, computer technology law, and many more diverse subsidies (2 percent), social assistance (1 percent), export tax breaks (Kandir Law) (1 percent), and other expenditures (11 percent). Only 6 percent of the expenditures were subject to being withheld by the executive to reach fiscal targets. However, the total value, which is approximately R$20 billion, is still significant. There has been much debate over what powers the constitution and the Fiscal Responsibility Law actually confer on the president in terms of discretion to execute the budget (Lima 2004). There has also been much debate on whether and how hard-wired items should be made flexible. As the rigidities imposed in the past on current public expenditures become more constraining and impede efficient adjustment to current circumstances, the pressure for decoupling some expenditures has increased. Also, as more expenditures become hard-wired, there are fewer resources left over for use as pork by the president for trades with congress. The only mechanism that remains largely unaffected by these fiscal trends is the political allocation of patronage to members of the governing coalition. But as indicated before, the increasing professionalization of the line ministries of the social sectors (the largest bureaucracies of the state machinery) has also restricted patronage to a certain extent. This has created an incentive for the president to try to find creative ways to increase his leeway for determining expenditures. The most frequent mechanism used by Presidents Cardoso and Lula has been â&#x20AC;&#x153;deconstitutionalization.â&#x20AC;? Hard-wired policies generally include those with little ideological content, and are perceived in society as entitlements. The two that best fit this description and that will be discussed in the next two subsections are education and health policies.

147

148

Lee J. Alston, Marcus AndrĂŠ Melo, Bernardo Mueller,

and

Carlos Pereira

Health Policy. The Constitution of 1988 created a unified budget for pensions, social assistance benefits, and health care, the so-called social security budget. This was part of the demand for a universal social protection system advanced by the opposition during the military regime and an important sectoral banner during the Constituent Assembly. A diversified source of funding was set up. This institutional arrangement was viewed by the groups supporting the idea as a mechanism that would delink contributions and access to the system, making it more democratic and redistributive. To this end, the constitution gave universal access to health care through the newly created Unified Health System (SUS). It also introduced generous social assistance benefits. The social security budget was made up of the CSLL, the contribution on net profits paid by corporations (Cofins), and the employersâ&#x20AC;&#x2122; and employeesâ&#x20AC;&#x2122; payroll contributions. The fusion of expenditures for health care and pensions in the same budget produced a dynamic over time that has been highly detrimental to health care, because pensions are contractual disbursements and are not compressible. They are a flux of future commitments that ends only with the death of the pensioners. By contrast, health expenditures are mostly current expenditures that are by definition vulnerable in the context of fiscal management. Over time, social security commitments crowded out health expenditures. It did not take long before this process became critical. This occurred because fiscal imbalances in the pension schemes were not very significant before the Constitution of 1988, and, more importantly, pensions were not indexed. This gradually resulted in a sharp reduction in the real values of pensions. By mandating that pensions were to keep their real value, the Constitution of 1988 resulted in future fiscal imbalances. It dramatically expanded the mass of workers under the civil service regime, upgraded rural noncontributory pensions and social benefits to the level of urban pensions, and set the lowest value of pensions at the minimum salary level. This produced a shock to the system and caused a crowding out of health expenditures shortly after its implementation. At the same time that the fiscal burden of pensions was growing, Brazil decentralized health care. The decentralization was very significant.

the

Road

Good Governance

The municípios were responsible for 9.6 percent of total spending in health care in 1985 (Arretche 2003). This share reached 35 percent in 1996 and 43 percent in 2000. The flip side of the coin is that the federal government’s share declined from 73 percent to 53 percent over the same period (Arretche 2003). The recurrent crisis in the funding of the health sector enhanced the visibility of health issues in the country. Brazil’s infant mortality rates are far above countries with comparable levels of per capita income. Policy elites have been increasingly sensitive to the need to address health issues as a precondition for development. Many proposals have been advanced for earmarking resources for the health sector. These have been criticized by the finance and planning circles as a move backward that would cause more fiscal rigidities in a context of rapidly declining degrees of freedom in the budget. The measures to secure financing for the health sector culminated in a proposal to create the CPMF, the provisional contribution on financial transactions. The CPMF was created by Constitutional Amendment 3 in 1993 and was a “sunset” provision that would be valid for only two years. In 1996, Constitutional Amendment 12 reinstated the CPMF and earmarked it for the health sector. In 2000, Constitutional Amendment 29 stipulated minimum values for investments in the health sector for the three tiers of government: municipal, state, and federal. For the federal government, the budget for 2000 was set at the 1999 level plus 5 percent. For the period 2001 to 2004, the value of health expenditures was readjusted by the annual variation of the nominal GDP. Of these, 15 percent was to be spent in the municipalities on basic health care, and distributed according to their population. In the case of the states, 12 percent of the revenue (after legal transfers to the municipalities) was to be spent in the health sector. In turn the municipalities were required to spend 15 percent of their budget on health care. The states and municipalities, which in 2000 had expenditures levels below those stipulated, were to reduce the difference at the rate of one-fifth per year. Noncompliance would allow federal intervention in subnational governments. The law stipulated that all transfers would be channeled to a fund and subject to auditing.

149

150

Lee J. Alston, Marcus André Melo, Bernardo Mueller,

and

Carlos Pereira

The initiatives aiming to hard-wire health care resources may be interpreted as attempts at controlling and securing subnational spending in a context of rapid decentralization and consequently high uncertainty over outcomes in an increasingly salient issue for the executive. It is significant that in the context of fiscal adjustment, discretionary health transfers—which are by far the largest of their kind in the country—have become crucial for the fiscal survival of subnational governments. This made the control of subnational spending all the more critical for the central executive. Education Policy. Education policy is another area that illustrates the use of institutional innovations as a precommitment device to ensure that they are preserved. As in the case of health care, primary (ensino fundamental) and secondary education are viewed by the governing coalition as crucial, thereby requiring insulation from political logrolls. Similarly, the key issue here for the executive is guaranteeing that the resources earmarked for primary and secondary education are in fact applied by the subnational governments in the sector and in specific ways. Article 30 of the Constitution states that primary education is to be provided by the municipalities, with the financial and technical assistance of the federal government and of the state. Constitutional Amendment 14 contains articles calling for priorities to be given by each level of government, but does not mandate specialization of competence. This definition provides an incentive structure that discourages efficiency because it diffuses responsibility and accountability. In the late 1980s and 1990s, the centrality of education to development became a recurrent issue in the public agenda. From business interests to social movements, a consensus emerged that was reflected in the executive’s preference for insulation of the educational sector. This was combined with the increasing need to enhance control mechanisms in the context of an accelerated program of decentralization. The furor over the quality of education finally resulted in the passage of Constitutional Amendment 14 in 1996 and the approval of the LDB (the complementary law of basic guidelines for education) in the same year. Constitutional Amendment 14 required that for 10 years, at least

the

Road

Good Governance

60 percent of the 25 percent of the subnational resources mandated for education were to be spent on the payment of teachers actively involved in classroom activities. The federal government was to play an equalization role. This has become the Achillesâ&#x20AC;&#x2122; heel of the new arrangements. Consistent with their preferences for fiscal expansion at the local level, subnational governments have pressured the federal government to raise its transfers. However, fiscal needs have led the national government to not readjust it. The Lula administration did not increase its funding, notwithstanding the intense criticism by the Workersâ&#x20AC;&#x2122; Party (PT) of inadequate funding by the Cardoso government. The federal governmentâ&#x20AC;&#x2122;s desire is to control subnational priorities and spending, while keeping its own preferred federal fiscal targets as the ultimate adjustment variable. The incentive structure led mayors to actively engage in attracting pupils because this would lead to more transfers from the fund. In addition, it encouraged decentralization from states to municipalities because there would be negative transfers in some municipalities if the educational services were provided by the states. The rationale for the education initiatives by the federal government is similar to those that underlie the health care system: attempts to control and secure subnational spending in a context of rapid decentralization and consequently high uncertainty over outcomes. Control by the federal government over subnational spending is therefore critical to ensure that the federal government meets its preferences. Conclusion

This study has found that the driving force behind policies in Brazil is the strong set of powers given to the president by the Constitution of 1988. To have strong powers does not mean unbridled powers. Several institutions constrain and check the power of the president, in particular, the legislature, the judiciary, the public prosecutors, the auditing office, the state governors, and the constitution itself. The electorate of Brazil holds the president accountable for economic growth, inflation, and unemployment. Because of the electoral connection, and perhaps because of reputational effects, presidents in Brazil have a strong incen-

151

152

Lee J. Alston, Marcus André Melo, Bernardo Mueller,

and

Carlos Pereira

tive to pursue stable fiscal and monetary policies as their first priority. At least for the past 10 years and especially during the Lula administration, executive power has been aimed at pushing policy towards macro orthodoxy. Although orthodoxy may not lead to short-term growth, international financial markets provide additional incentives for discipline, as deviations are instantly punished with unfavorable consequences that are readily recognized by the electorate. Achieving stable macro policies required constitutional amendments as well as considerable legislation. To attain their goals, past administrations (Cardoso and Lula in particular) used their property rights to pork to trade for policy changes. The rationale for members of congress to exchange votes on policy for pork is that the electorates reward or punish members of congress based on the degree to which pork lands in their district. With the exception of the devaluation of 1999, macro policy has become more stable over time. Macro policies in Brazil can be characterized as “stable but adaptable.” The pursuit of macro orthodoxy comes at a cost; some policies in Brazil are “volatile and unstable.” This analysis has found volatility and instability in policies that have an ideological component, such as land reform and poverty alleviation, or whose gains accrue at the congressional district level, such as infrastructure projects. The volatility on ideological policies is no surprise and happens most when administrations change. For policies having a local rather than national impact, the volatility results from the fact that spending is based on the residual left in the budget after the president takes care of hard-wired and pork expenditures. The negative side of this arrangement is that many infrastructure projects have fallen into this residual category, notably sanitation and local roads. The Constitution of 1988 is a detailed political document rather than a set of principles. It has constrained and still constrains policymaking. Its biggest constraint comes from hard-wiring expenditures in certain policy areas, most importantly those in health and education. The constraints from the constitution bind because the judiciary has been relatively independent in ruling on issues of constitutionality. This has led to a perception that the constitutional amendments will be permanent. Reinforcing the role of the judiciary is the independent and increasingly

the

Road

Good Governance

powerful role played by the Ministério Público. Public prosecutors have been most active in enforcing social policies, such as legislation concerning education, health, and the environment. Despite the perception in the press, the power of governors to influence national policies is relatively weak and is becoming weaker. In large part this resulted from the enforcement of a law on fiscal responsibility. State and local political actors still play a role that affects economic performance. Considerable federal and state funds are diverted from public spending intended for public goods and infrastructure toward campaign finance. Future research by the authors will investigate the role of the courts and auditors’ offices in allowing corruption (the misuse of public funds for private gain) to persist. Overall, this study has painted a relatively rosy picture of the policymaking process in Brazil. Importantly, pork was found to be a relatively cheap and effective means to ensure stable but adaptable macro policy. The actions to date of the Lula administration bolster this view. Of course there are deficiencies: most prominently in the inflexibility of health and education policies, the volatility of social programs and infrastructure investments, and the persistence of corruption. But social programs can be advanced in a permanent fashion only when the government has ensured stable macro policy and the confidence of world capital markets. Brazil is currently in a relatively stable political equilibrium, but this equilibrium can be upset by a sufficiently large exogenous shock.19 A shock of sufficient magnitude could tip Brazil back to its former populist ways.

For a discussion of Brazil’s use of fiscal and monetary policy as an equilibrating mechanism in response to a monetary shock (the 1999 devaluation) and a political shock (the election of Lula), see Alston and others (2007). The actions taken in response to those shocks were consistent with the framework presented in this chapter. 19

153

5 Political Institutions, Policymaking Processes, and Policy Outcomes in Chile Cristóbal Aninat, John Londregan, Patricio Navia, and Joaquín Vial

Introduction

This chapter focuses on the policymaking process (PMP) in Chile since 1990, when the military dictator Augusto Pinochet turned over the presidency to his democratically elected successor, Patricio Aylwin. That moment represents a sharp change in Chile’s institutional framework, so it makes sense to treat the ensuing period as a distinct policymaking regime. Since the restoration of democracy, Chile’s institutional system of checks and balances has worked well in terms of promoting intertemporal political transactions. Chile has a presidential system with a bicameral congress, a proportional electoral system, an independent judiciary, and other enforcement technologies. The system produces a PMP favorable to intertemporal political transactions in most policy areas, as it favors a small number of political actors who interact repeatedly with long time horizons, and there are good enforcement technologies overseeing the process. Four salient institutional features together have shaped policymaking in Chile during this period. First, there is the party system, which consists of two closely knit and stable coalitions, one on the left and the other on the right. This party system is strongly shaped by the electoral system, which creates pressures for moderate polarization in legislative elections through the use of the d’Hondt rule with open-party lists and a district

156

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

magnitude of two. A “dual ballot” system for electing the executive discourages extremist presidential candidates.1 The second salient feature of the PMP is a powerful agenda-setting executive. The president of the republic has near-monopoly control over the legislative agenda, with proposal and veto powers that make him a de facto agenda setter. Third, the PMP is studded with veto players, some of them written into the constitution by the outgoing military government in order to make policy changes by subsequent elected governments more difficult. These include a bicameral congress, a comptroller general, and independent loci of judicial power, including the regular courts, a constitutional tribunal, and an electoral tribunal. Several significant veto points were removed by a constitutional reform in 2005, including a provision for unelected senators in the upper chamber of congress and the relative autonomy of the armed forces (the heads of the armed forces could not be removed by the president of the republic, and they were a dominating presence on the National Security Council, or COSENA).2 A fourth key feature of the Chilean PMP is the existence of a wellfunctioning mechanism for policy implementation, including an independent judiciary and an honest and reasonably efficient bureaucracy (although there is variation across policy areas on this last score). As discussed further in the chapter, because the party system shrinks the de facto set of decision makers and because the parties are long lived, the Chilean system is characterized by a relatively small set of decision makers who repeatedly interact. The relatively honest and well-working bureaucracy and judiciary facilitate cooperation by enhancing transparency (through adherence to standardized procedures), and providing reliable enforcement of the policies that are put in place. The existence

In a dual ballot system, if no candidate obtains an absolute majority of the popular vote, there is a second round between the candidates who obtained the two highest vote shares. 2 These undemocratic institutions were introduced by Pinochet in the 1980 Constitution. After 15 years of negotiations between the center-left government coalition and the right-wing opposition coalition, they were eliminated by a constitutional reform in 2005. 1

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

of a transparent enforcement technology also limits the benefits obtainable by defecting from agreed-upon policies. The degree to which policymakers’ interests align differs across issues. However, two features of the Chilean polity are worthy of particular note. First, Chile has a small open economy, so that the consequences of poor economic policy are particularly keenly felt. Second, there is the “hangover” from the civil strife of the early 1970s and from more than a decade and a half of military rule; certain sensitive issues create a shared interest in avoiding conflict. Beyond these two features, the alignment of interests among policymakers varies from issue to issue. The framework for this chapter is based on work by Spiller and Tommasi (2003) intended to identify features of the PMP that tend to produce cooperative outcomes. Cooperation means working together for a common purpose or benefit, and this can happen only when there is at least some alignment of interest among the participants, and when the degree of alignment between the parties places a ceiling on the amount of cooperation that one can expect to observe. The Spiller and Tommasi (2003) framework emphasizes five factors identified in the game theoretic industrial economics literature as facilitating cooperative outcomes:3 first, there is only a small number of decision makers; second, there is repeated interaction among decision makers; third, deviations from agreed-upon behavior are easily observed; fourth, the immediate benefits from reneging on agreements are small; fifth, credible enforcement mechanisms exist. Left off their list is the degree to which agents’ interests align, though the alignment of interests is central in the industrial organization literature from which the list is derived. Accordingly, we take the advice of Spiller, Stein, and Tommasi (2003, p. 6, footnote 8) to combine the “lens with which authors … can focus the analysis of the PMP” with “others that they believe relevant for understanding key features of their respective countries,” and add a sixth factor to the list of those facilitating cooperative outcomes: sixth, the degree to which the parties’ interests are shared. While all six of these factors play an For more on this literature, see Friedman (1971); Green and Porter (1984); Abreu (1986). 3

157

158

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

important role in the Chilean policymaking process, the first, second, and sixth factors are particularly salient. An additional consideration bearing on the repeated games literature from which the unifying theoretical framework of this chapter springs is that the factors it identifies as facilitating cooperation are factors that expand the set of possible equilibria to include cooperative outcomes, along with “uncooperative” myopic equilibria. That is, in the parlance of game theory, the oligopoly models that allow for possible cooperation also have other equilibria consisting of endless price wars. The game-theoretic models provide little guidance as to which of the possible equilibria will actually occur. Thus one may think of the factors identified by the framework applied here as identifying an upper bound on the level of cooperation that can emerge in a given political system. It may be that some significant differences between, say, Uruguay and Chile—both of which have relatively small and stable party systems and independent courts—result from policymakers having coordinated on different equilibria among the set that are supported by their institutional framework. Thus one may think of the six elements identified by the preceding discussion as “risk factors” for the emergence of cooperation. While the presence of these factors does not guarantee that political actors will successfully coordinate on one of the cooperative equilibria, it does at least hold out the possibility that they will do so. The second section describes Chile’s policy outcomes and the policymaking capabilities of the Chilean state from a comparative perspective. It focuses on the outer features of policies following the criteria set by Scartascini and Olivera (2003). The third section describes and analyzes Chile’s institutional setting and the incentives it sets for political actors. First, it analyzes the combined effects of the two different electoral systems, for presidential and legislative elections, and the congressional rules over the party system and the number of political actors. The subsections that follow describe the powers of the executive and its role in the legislative process, review the role of enforcement technologies and the civil service on policy implementation, and analyze the interactions of the different actors in the policymaking process. The fourth section concludes.

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

Policy Outputs and Policymaking Capabilities Stability and Flexibility

Public policy in Chile has shown a remarkable level of stability since the last years of the dictatorship, in the face of major political changes entailed by the transition to democracy and three later changes of government. The four different Concertación administrations that have governed the country since 1990 have maintained fiscal discipline and avoided manipulation of fiscal policy for political purposes. A fiscal surplus was consistently maintained between 1990 and 1998. During the Asian Crisis, recession pushed the budget into a deficit of 1.4 percent of GDP in 1999, but the budget was again balanced the following year. The government quickly addressed the problem and proposed the adoption of a new fiscal rule in March 2000. The new rule seeks a “structural” surplus of 1 percent of GDP. It was designed to prevent inefficiencies produced by erratic and discretionary government spending, maintain stability in fiscal policy, and increase transparency in government spending.4 However, unlike other public policy features, the “1 percent fiscal surplus rule” is a discretionary policy self-imposed by the last two administrations and might not survive beyond the end of Bachelet’s term in March 2010. High rates of GDP growth during the 1990s allowed governments to increase public expenditure while repaying most foreign debt, thus assuring a solid macroeconomic position to foreign investors. The ratio of government debt to GDP decreased from 42.7 percent of GDP in 1990 to 13.3 percent in 2003 (DIPRES 2004). The relatively conservative fiscal response of the Chilean government to the “boom” years of the 1990s can be examined in the context of Latin American countries’ tendency to pursue procyclical fiscal policy, and runs counter to the stereotype of developing country governments trapped in “boom and bust” cycles. Monetary policy has also been conservative since the restoration of democracy. The central bank applied restrictive monetary policies dur4

For more details regarding the structural balance rule, see Marcel and others (2001).

159

160

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

ing most of the 1990s, achieving a gradual reduction in average inflation rates from 26 percent in 1990 to 1.1 percent in 2004.5 Between 1989 and 2000, the central bank worked with one-year inflation forecasts to lower expectations of future inflation. In 1999, the central bank announced a policy shift, effective in 2001, to an inflation-targeting framework (IT). The IT reduces the central bank’s discretion and puts automatic stabilization mechanisms in place to maintain inflation within the target band of 2 to 4 percent. This simple targeting rule tends to protect the central bank from political pressure. The IT framework for monetary policy has been successful, as inflation has remained within the band since the inception of the system in 2001. Chile today enjoys a stable and very low level of price inflation, and its fiscal and monetary institutions have earned praise from multilateral institutions, credit rating agencies, and international investors. Trade policy is another example of stable policymaking with successful results. The military dictatorship initiated an aggressive unilateral liberalization strategy in the mid-1970s to eliminate the high trade barriers that had protected domestic economic production for most of the twentieth century. The Concertación governments continued with the unilateral liberalization policy. Tariff rates averaged 11 percent during the 1990s. Then a four-year-long series of tariff reductions beginning in 1999 lowered across-the-board tariffs to 6 percent in 2003. The Concertación administrations complemented the unilateral liberalization policy with bilateral economic agreements during the first half of the 1990s with several Latin American countries, including the Mercosur trading bloc. Since 1996, Chile has signed free trade agreements with Canada, the European Union, the United States, the Republic of Korea, and other countries. Today Chile has one of the freest economies in the world, trade volumes closely follow the evolution of terms of trade, and the share of trade in GDP has risen steadily from 28 percent of GDP in the 1960s and early 1970s to more than 70 percent today. In September 1999, the central bank decided to abandon an exchange rate policy based on a price band and instead adopted a free5

Central Bank of Chile, www.bcentral.cl.

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

Figure 5.1. Trade Liberalization, Chile, 1974–2004 108

80 75

Percentage of GDP

40 35

Percentage of GDP

30 12

0 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Year Trade openness1 (left axis)

Average tariff (right axis)

Source: Authors' calculations based on “Base de Datos Estadisticos,” Central Bank of Chile. 1 Trade openness is calculated as the share of total trade (exports and imports) in GDP.

floating exchange rate policy. The floating exchange rate policy gives the economy enough flexibility to face external shocks and reduces the discretionary power of the central bank. The central bank has twice intervened in the exchange market in recent years: in July 2001 because of the crisis in Argentina, and in October 2002 because of a crisis in Brazil. However, on both occasions, the interventions have been transparent and well founded, with the central bank providing a public rationale for its actions and providing full disclosure as to the amounts involved in the operations and the timeframe for the unusual action. Cross-national indexes of economic policy also reflect the stability of Chile’s policy outcomes. Since its transition to democracy, Chile has improved consistently in indexes such as the Heritage Foundation’s Index of Economic Freedom, the World Economic Forum’s Growth Competi-

161

162

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

tiveness Index, Transparency International’s Corruption Perceptions Index, and the World Bank’s Governance Indicators.6 Coordination and Coherence

Government policy has been coherent across policy areas. In part this suggests a long-term strategy for economic policy, and in part it is the result of a shared commitment by both the government and the opposition coalitions (the Concertación and the Alianza, respectively) to promote efficient economic policy. During these past years, changes made in key policy areas—including the structural balance rule, a monetary policy with inflation targets, and a free-floating exchange rate—have made the entire public policy process more transparent, coherent, and internally consistent. In a country with a flexible exchange rate like Chile, the inefficacies and inefficiencies of fiscal policy counsel against using them as instruments to bring about stabilization. Instead, their main role should be to secure a solid fiscal position. That better serves as a basis for a stable monetary policy and serves to consolidate long-term economic growth. In this way, the fiscal rule of government surplus adopted in 2001 under the tenure of Finance Minister Nicolás Eyzaguirre is a significant innovation, as it acts as an automatic stabilization tool. Similarly, since the free-floating exchange rate was adopted in 1999, a potential source of incoherence in the formation of public policy has been eliminated. Before that decision, the central bank had to worry about keeping the exchange rate within the target band at the same time that it sought to control inflation, which occasionally produced competing and contradicting objectives. The costs of these tensions between competing objectives became evident in 1999, when the central bank was forced to raise interest rates well beyond prudent limits when the exchange rate band came under pressure after the Russian debt default. The only central bank objective now is to meet inflation target goals. See Kaufmann, Kraay, and Mastruzzi (2003); Miles, Feulner, and O’Grady (2003); Transparency International (2003); World Economic Forum (2003). 6

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

While the military government left what Navia and Velasco (2003) refer to as a â&#x20AC;&#x153;first generation of reformsâ&#x20AC;? in place, subsequent democratically elected governments continued the reform process with what have been called second-generation reforms, including privatization of the port sector and utilities, introduction of private concessions in public works, and educational, health, and labor reforms. These reforms, however, have been achieved with varying degrees of success. The most advanced areas correspond to macroeconomic policy, where there is considerable consensus. Also, substantial success has been achieved in improving the regulatory framework, the transparency of government policies, and state capabilities. Areas where success has been more elusive are those related to human rights, social issues (such as divorce), and efforts at reforming the labor code and the import-competing agricultural sector (wheat, vegetable oils, and sugar). The introduction of public works concession programs has been a successful innovation. Historically, the state was in charge of building and operating public infrastructure works. In the 1990s, however, the state initiated an aggressive concessions plan through a system of Build, Operate, and Transfer contracts (BOT). The use of such a system brings about huge improvements in efficiency and welfare. From a distributive approach, it is convenient that the users of the new infrastructure pay for it, rather than placing the burden on all taxpayers. Finally, the design of the contract seeks to generate competition and helps discard ideas where the expected demand does not justify the construction of the new infrastructure (MOPTT 2003). There are also a large number of public policies that work in a manner directly opposed to the above-mentioned concessions program: they carry very heavy political costs in the short term, but yield enormous long-term benefits. If those political property rights are not protected, it will be difficult for governments to undertake those kinds of reforms. Two sectors that clearly present these kinds of challenges are health and education. Not surprisingly, progress in these areas has been far slower than in concessions of public works. In the education sector, spending has increased by more than 220 percent between 1990 and 2003 (DIPRES 2004). That trend has allowed

163

164

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

the government to reverse the decline in educational spending observed in the 1980s. As a result of a correct diagnosis of the bad state of Chilean education, an educational reform (elementary and secondary) was adopted in the 1990s. A number of initiatives to improve the system were undertaken: improvements in working conditions and salaries for teachers, improvements in structure and equipment for existing schools, improvements in access to education (reflected in an increase in enrollment figures), and the most comprehensive curriculum reform in 25 years. Together with improving the quality of education, enormous efforts have been made to reduce school dropout rates. Yet much remains to be done. Access to higher education remains highly discriminatory against low-income youth. Because the public funding scheme for university education covers only public institutions, many capable young students who attend private universities are left without access to public funding for their university education. In summary, although some important reforms have been undertaken in recent years with the objective of giving more autonomy to educational institutions, linking teachers’ pay with their performance, providing more information to parents on school performance, and improving access to higher education to all students, the successes in this field have been positive but largely insufficient. In the health sector, there are some signs of progress, but some worrying signs of neglect as well. In the early 1990s, there were significant deficiencies in infrastructure and equipment in the public health system, a lack of qualified human resources, and administrative inefficiencies. To correct those problems, the government invested more in the public health system than it had at any previous time in the country’s history. Health spending went from 1.9 percent of GDP in 1990 to 3.0 percent in 2003. The increase is also evident when comparing health spending as a share of total expenditures; it went from 9 percent of total expenditures to 14 percent during the same period (DIPRES 2004). Even though Chile has achieved health indicators (lifetime expectancy at the time of birth, infant mortality, and mother’s mortality) higher than what its economic development level would predict, public satisfaction with health coverage remains low. One of the main criticisms

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

against the system is that it does not constitute a system per se. Instead, there are two systems (one public, one private) that function in parallel, with little coordination and insufficient cooperation between them. A partial health reform guaranteeing speedy access to the public system in case of a limited number of illnesses was approved in 2004, but it falls far short of what is needed to improve the efficiency and coordination of the whole system. Opposition by the Alianza and by other interested parties—including doctors’ and health workers’ unions, with the tacit or explicit support of members of congress belonging to the government coalition—was able to derail important structural changes initially proposed by the Concertación administration.7 Despite repeated efforts to improve labor markets, success has been moderate. The first labor reforms adopted after the restoration of democracy (1990 and 1991) sought to correct the imbalance between workers and employers. That imbalance was inherited from the labor reforms adopted by the Pinochet dictatorship during the 1970s and 1980s. The lack of legitimacy of the labor system inherited from the dictatorship weakened its long-term stability. Thus while the new democratic government sought to maintain some of the improvements made in the 1970s and 1980s regarding flexibility and modernization, other rules on individual contracts, collective bargaining, and the operations of labor unions were introduced to bring legitimacy to the system and to bring labor unions on board as allies of labor market reform efforts, rather than enemies (Mizala and Romaguera 2001). Yet during the last year of the Frei administration, a labor reform was launched in an effort to reverse the low levels of labor union participation and collective bargaining by seeking to bring some additional protection to workers and labor unions. The labor reform initiative was rejected by the legislature in the midst of the 1999 presidential campaign, to no one’s surprise. Yet President Lagos, after taking office, sent

The former chairman of the doctors’ union is now a member of congress representing the PPD (the president’s party), and was very active in the health commission of the chamber of deputies, which is almost completely monopolized by medical doctors. 7

165

166

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

a new labor legislative initiative to congress that eventually passed. The new legislation had two objectives. First, it sought to increase the legal protection of workers (by formally recognizing a number of rights that were not clearly spelled out in the existing legislation). Second, it sought to reduce the costs of hiring new labor. These dual—and somewhat contradictory—goals have made it difficult to appropriately evaluate the success of the reform. A significant improvement in labor market regulation was achieved in 2002. A mandatory unemployment insurance scheme for those who enter the labor market during or after 2002 was passed into law after some tough negotiation and bargaining with the opposition and left-wing Concertación legislators. Eventually, the newly created unemployment insurance scheme should replace the existing dismissal compensation scheme. A novel feature of this system is the introduction of individual savings accounts with contributions by both the workers themselves and companies to fund compensation during the unemployment period. However, the law did not include a revision of the high dismissal compensation scheme, which puts a very heavy burden on companies that seek to lay off workers and introduces severe distortions into the Chilean labor market.8 According to Lora’s structural reform index for the 1985–2001 period (Lora 2001), reforms in Chile have increased the efficiency of the public policies and have strengthened free markets for the five economic sectors considered. The level of progress and liberalization of each of those five sectors is different, but there is no incoherence in the reforms implemented in each one of those sectors. The most important advances took place in the commercial and financial arenas. The labor index is the only indicator that arguably showed a decline toward the end of the 1990s. Yet as discussed in the previous section, the labor reforms implemented during the 1990s sought to restore certain workers’ rights that were lost in the previous decade.

For more information on this topic, see Cowan and others (2003).

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

Public Regardedness of Public Policy

Government spending in Chile by and large reaches those in most need. The last poll on poverty and inequality, CASEN 2000,9 shows that fiscal spending on health, education, and direct subsidies for the poor was well spent; 68.8 percent of all spending on these items went to the poorest 40 percent, and an impressive 80.2 percent of public spending in health benefited the poorest 40 percent of the population. In terms of education, 62.8 percent of spending went to the poorest 40 percent of the population. In terms of direct subsidies, 73.1 percent of all that was spent went to the poorest 40 percent of the population. Altogether, the subsidies and benefits that the state offers to the poorest Chileans allowed the first decile (the poorest 10 percent) to increase its share of income from 3.7 percent before taxes and subsidies to 6.4 percent of total pre-tax income, while the wealthiest 20 percent saw its share of income reduced from 57.5 percent before subsidies to 53.4 percent after subsidies. The distribution of social transfers per quintile in Chile as well as nine other Latin American countries is presented in figure 5.2. The figure shows a stark contrast between the focus of social spending in Chile and in other countries of the region. While in Chile transfers steadily decline from the poorest quintile (quintile I) to the richest one (quintile V), in the other countries, without exception, transfers tend to favor the rich. However, there are some areas where highly intense private interests have captured the policymaking process. A notable example of this is the price bands for some agricultural goods. Chile introduced price bands for some agricultural products with the aim of reducing tensions in politically sensitive southern regions of the country. The price bands are applied to wheat, sugar, and vegetable oil. The argument behind the decision was that highly fluctuating international prices for these goods placed an overwhelming burden on producers. A protectionist neutral band was adopted so that the government could impose tariffs when international prices were too low and reduce tariffs when international 9

See MIDEPLAN (2000).

167

CristĂłbal Aninat, John Londregan, Patricio Navia,

and

JoaquĂn Vial

Figure 5.2. Distribution of Social Transfers by Income Quintile, 1997â&#x20AC;&#x201C;2001a 70 60 50

Percent

168

40 30 20 10 0 1

Quintiles Argentina

Bolivia

Brazil

Chile

Colombia

Costa Rica

Mexico

Peru

Uruguay

Venezuela

Source: de Ferranti, Perry, Ferreira, and Walton (2004). a Excluding pensions.

prices were too high. The scheme was not deemed protectionist, but rather was defended on the grounds that it would reduce price volatility, thus improving the welfare of agriculture workers. Yet this mechanism has not functioned correctly. Lobbying by agricultural producers has prevented the government from reducing tariffs when international prices have increased and has even effectively altered the scheme by which the government calculates the appropriate price bands for these products. The benefits from price intervention go primarily to well-off agricultural producers, not to minimum-wage agricultural workers. Galetovic (2001) showed that 63 percent of small sugar beet producers produce only 4 percent of the national production and receive less than 1 percent of the total amount overpaid by consumers for imported sugar (US$258

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

million in the 1998–2000 period). Taking into account its effect on consumers and producers, this price intervention policy is probably one of the most regressive policies that exist today in Chile. State Capabilities and Overall Quality of Public Policy

Chile’s policy schemes and regulation are relatively transparent. There is a simple tax code, tax evasion is low, and rates of effective and total trade protection are very similar. The procedures and time required to open new businesses and to comply with government regulations are much lower than regional averages (but are still higher than in OECD countries). The regulatory burden is moderate. While survey data reveal that the time and government fees needed to comply with regulations when a new business is opened are low by international standards, the total number of regulations that must be complied with exceeds the international average. The staff of the regulatory agencies, tax collection agencies, central bank, and finance ministry tend to be more prepared, better trained, and more professional than that of other government agencies. It is well known that the liberalization of the economy and of the financial system, the privatization of companies that offer public services, and the consolidation of markets must be accompanied by the strengthening of the institutions in charge of regulating those activities. Otherwise, there will be high risks of capture where the regulator works to advance the interests of the regulated industries rather than to defend the interest of consumers and to secure more competitive markets. The option of choice for Chile’s regulatory scheme has been based on reducing entry barriers and on selective, rather than structural, intervention executed by antimonopoly commissions that bring more flexibility to the system. It would be far-fetched, however, to suggest that regulatory policies have changed since 1990. The so-called “rules of the game” have not undergone significant changes. Instead, there has been an effort to improve the regulatory framework in light of specific problems that have arisen over time. With regard to the quality of the personnel who work for the central government, there is a high degree of centralization of relevant deci-

169

170

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

sions. There is a limited margin for decision making at the regional and local institutional levels, and there is a high level of inflexibility, with little room for discretion for those responsible for the management of the respective institutions and offices. Moreover, there is a single and uniform regime of human resources for the entire public sector. That makes it more difficult to adjust to the specific needs and demands of particular institutions and public services. There is a very rigid employment stability framework, which in practice amounts to an inability to fire public workers. Salary and public compensation also respond to a rigid and nationally controlled scheme that often favors those who have been employed by the public sector the longest. Since 1994, there have been several attempts to modernize the state bureaucracy. The most remarkable was the creation in 2003 of a public civil service division (Nuevo Trato y Dirección Pública). This new office will fill between 1,900 and 3,500 positions that were previously considered presidential-appointed positions. Two autonomous institutions were created to appoint the new public sector employees for those positions: the Consejo de Alta Dirección Pública and the Dirección Nacional del Servicio Civil. Other institutional changes of note adopted in Chile since 2003 are a new law on government remuneration and spending that regulates salaries for high-level officials and caps their discretionary budgets; new legislation that requires the government to maintain a registry of all individuals, institutions, and companies that receive public funds; and a new law on political party and campaign financing that has brought the issue of money and politics to the forefront of debate in Chile. Most recently, the legislature also began to debate a proposal to regulate lobbying. Among the noteworthy features of that legislative package are the proposal to register and identify all official lobbying organizations and individuals as well as meetings between lobbyists and public authorities, the provision of equal access to authorities for all lobbyists, restrictions on public sector employment and electoral eligibility for those who work as lobbyists, and restrictions on political contributions by lobbyists. Lastly, indexes that measure corruption, governance, business environment, and quality of institutions are useful proxies to assess the over-

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

all quality of public policies from a comparative perspective. The “Doing Business” Index prepared by the World Bank ranks Chile above Latin American countries, but well below OECD averages.10 This index assesses regulations and policies insofar as they foster (or hinder) investment, productivity, and growth. Some of the regulations in Chile include the following. In order to start a business in Chile, a person must complete 10 steps (procedures). That is below the Latin American average, but is significantly higher than in OECD countries. In order to enforce a contract, 21 procedures are needed in Chile, requiring an average time of 200 days and costing an expected 14.7 percent of GDP per capita. Although those values are also lower than the Latin America average, the cost of enforcing contracts is much higher than in OECD countries, where it stands at 7.1 percent of GDP per capita. Chilean courts normally take 5.8 years to rule on insolvencies, much longer than in OECD countries, where courts take on average 1.8 years. The Corruption Perception Index, produced by Transparency International, ranks Chile among the 25 least corrupt countries in the world, and as the least corrupt country in Latin America.11 According to the Global Competitiveness Report, Chile ranks 26th among 123 countries in the world.12 This report, prepared by the World Economic Forum, analyzes three fundamental pillars for growth: business environment, good institutions, and technological development. Chile’s highest ranking comes from the quality of its institutions. Political Institutions and the PMP in Chile

This section argues that Chile has an institutional system of checks and balances—with some similarities to the United States, but with some important differences—that has worked well in terms of promoting intertemporal political transactions since democracy was restored in 1990. Chile has a presidential system with a bicameral congress, a proportional See World Bank (2004a). See Transparency International (2003). 12 See World Economic Forum (2003). 10 11

171

172

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

electoral system with a district magnitude of two, an independent judiciary, and other enforcement technologies. The ensuing PMP is conducive to political cooperation in most policy areas, as it favors a small number of political actors who interact repeatedly with long time horizons, and there are good enforcement technologies overseeing the process. Some “undemocratic enclaves,” established in the 1980 Constitution left behind by the Pinochet regime (eliminated through a constitutional reform in October 2005), reduced the “issue space” for policymaking during the period under analysis, restricting cooperation in areas regarding human rights and some military-related issues. After describing the main components of Chile’s political system and analyzing the incentives they establish for political actors, the section explores their interactions and the combined effect on the policymaking process. Chile has a presidential system similar to that of the United States, with a bicameral congress as part of a system of checks and balances. Such systems are designed to allow for slow but incremental change, preventing policy instability as a result of changes in the balance of political power. If these systems are well designed, however, they also allow for decisiveness when there are external shocks that change the relative value of political issues for the relevant actors. The crucial issue here is not the existence of institutions designed to be checks and balances, but the precise design that determines the workings of those institutions. Several comparative studies note that policies are of very different quality in countries with similar institutional settings. The effectiveness of the institutional design is in its details; these will determine the possibility of finding loopholes or bypassing the checks and balances—which ultimately will determine the real incentives for political actors. Systems of checks and balances are designed to produce good policies in areas of relative consensus or in cases where compensation could possibly be worked out; to produce a bias toward the status quo in policy issues that are more conflictive; and to produce relative gridlock on issues where political actors have strong opposing views. In that regard, the Chilean system seems to be well designed. It has allowed for a constant and incremental improvement of policies, institutions, and rules

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

and regulations since the restoration of democracy. It has allowed for stability with respect to political shocks and for decisiveness with respect to socioeconomic shocks. As expected in a system of checks and balances, the quality of policies varies across policy areas. It tends to be higher with regard to macroeconomic issues, international integration, financial markets, and regulations, as opposed to social issues like labor markets, social security, health, and education. On moral issues, human rights, and military-related issues, policies show higher rigidity and overall lower quality. Effects of the Electoral System, Electoral Cycle, and Congressional Rules on the Party System

Chile has two different electoral systems for legislative and presidential elections, which give different incentives to political actors and crosspressure them in their electoral strategies. For legislative elections, Chile has the so-called binomial electoral system, which is a proportional representation system with a district magnitude of two in all districts. It uses open lists (with an upper limit of two candidates per list) and the d’Hondt seat allocation formula. It allows for coalition formation, but coalitions are nationally binding.13 In practice, each of the lists (coalition/parties) receiving the two highest vote shares wins one of the two available seats per district—unless the list receiving the most votes outpolls its second place rival by a ratio of more than two to one, in which case it receives both seats. Chile’s electoral system has a twofold effect on the party system: it reduces the number of relevant actors by encouraging parties to coalesce, and it strengthens the national leadership of parties. At a district level, a magnitude of two determines an upper bound of three parties (Cox

When a coalition has more than two parties, the provision that limits the number of candidates to be presented in each district by each list forces them to negotiate the districts in which each party presents a candidate. By the end of the period under study, this provision has led to high tensions in the Concertación, stressing their capacity for cooperation. 13

173

174

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

1997).14 At the national level there can be many parties presenting candidates in just a few districts; hence the system allows for the existence of several parties. But at the same time, there is a strong incentive to coalesce at the district level, since if the list receiving the most votes doubles the second place rival, it obtains 100 percent of the seats available in the district. If a list secures second place and avoids being doubled, it gets 50 percent of the seats being contested. The provision that coalitions are binding at a national level leads parties to form broad-based national coalitions. Since it is difficult to form coalitions that can secure more than two-thirds of the votes in each district, and it is relatively easy to secure one-third of the vote share, the most likely outcome is the formation of two national coalitions, which in most districts obtain one of the two seats available. Under this scenario, if coalitions are already formed, any given party would pay a high toll if it unilaterally left its coalition. On the other hand, this sets a high entry barrier for a third national coalition or independent party. Since 1990 there have been six parties with congressional representation, organized into two national coalitions. The 1988 plebiscite to decide whether General Pinochet should remain in power for eight more years organized the existing political actors into two blocs supporting the “yes” and “no” votes. These blocs have remained during the subsequent democratic elections, with the center and left-wing parties that opposed Pinochet in one coalition, and the right-wing parties that supported the general in the other: the Concertación de Partidos por la Democracia and the Alianza por Chile, respectively.15 The Concertación and the Alianza are

The district magnitude (DM) determines the minimum vote share required to secure a seat in each district. (For example, when DM equals two, a candidate can secure a seat with about one-third of the votes. The exact percentage depends on the electoral formula being used.) Regardless of the initial number of parties, in the absence of other influences, parties tend to consolidate in fewer and larger parties in order to maximize their chances to gain congressional seats. In a repeated game—as periodic elections are—this process continues until the number of parties equals DM + 1. 15 The Concertación originally consisted of 17 parties that opposed the Pinochet regime. During the 1990s, these parties merged or disappeared. The Concertación currently consists of the centrist Christian Democratic Party (PDC), the leftist Party for Democracy (PPD), the Socialist Party (PS), and the Radical Social Democratic Party 14

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

still the only two coalitions with congressional representation, and have become the longest-standing coalitions in Chile’s long history as a republic. The existence of these coalitions provides political parties with several incentives and restrictions. Since only one-third of the vote is needed to secure 50 percent of the seats in each district, it drives parties’ platforms away from the median voter, potentially polarizing the party system. (As indicated below, this effect is mediated by the effects of the presidential electoral system, which reduces its centrifugal incentives.) Also, the potentially high price a party would pay in terms of its share of the national vote if it unilaterally leaves its coalition enforces high intracoalition discipline. This has been important for maintaining unity in both coalitions throughout the democratic period, since they each include parties with different platforms on several issues, as well as political leaders with deep personal rivalries. Despite publicized intracoalition bickering between parties, both coalitions have remained united by the pressure of their congress members on their respective party’s leaderships. Congress members know that their chances for reelection would be jeopardized should their coalitions break up. The binomial system also encourages continuous intracoalition party negotiations to decide which candidates will be nominated to the coalition’s lists in every district. This strengthens the national leadership of parties, but only to a certain extent. In high-magnitude proportional representation systems, candidates must respond to their party leadership to be included in the ballot. In single-member districts, politicians are political entrepreneurs since they must respond to their local constituencies. Under the binomial system, politicians and parties have mixed incentives. The national party leadership must negotiate with its partners as to the districts in which it should present candidates while responding to both national and local considerations: on the one hand, parties must negotiate their share of the coalition’s candidate list, and on

(PRSD). Since 1989, the Alianza has had several names, but it has always consisted of two parties: the moderate right-wing National Renovation (RN), and the more rightist Independent Democratic Union (UDI).

175

176

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

the other, they must try to get the districts where the party has candidates with strong local support.16 The legislative electoral system does not have term limit provisions, which, together with the need for candidates with strong local support, encourage politicians to seek long legislative careers. In Chile 75 percent of congress members are renominated and about 60 percent are reelected (Carey 2002a). This reelection rate is much higher than in other Latin American countries, but it is lower than in the United States, where 99 percent of representatives seek reelection and about 80 percent succeed. (This issue is examined further in the section on the role of seniority in congress.) Summing up, the main effects of the binomial system for the workings of the policymaking process are that it reduces the number of relevant actors to a few parties organized into two broad-based stable coalitions, and strengthens the party leadership while encouraging politicians to respond to their constituencies to a certain extent, and to have long legislative careers. Finally, given the difficulty of one coalition doubling the vote share of the other in any given district, under the binominal system congressional representation for each coalition oscillates at around 50 percent of the members of each chamber of congress.17 For presidential elections, there is a majoritarian electoral system with runoff provisions if no candidate obtains a majority of the votes Even though the electoral system is the main determinant of the party system, the structure and behavior of political parties is determined by the incentives and restrictions established by the relevant institutional setting as a whole. The Political Parties Law, campaign finance laws, and internal rules of congress, among other factors, also influence political parties’ behavior. 17 In addition to the positive effects the binomial system produces for the workings of the PMP, it produces a negative effect on the degree of competitiveness of the party system, and its responsiveness to changes in the electorate’s preferences. Given the difficulty of one coalition doubling the vote share of the other in any given district, the binomial system leads to a situation in which each one of the two main coalitions has a secure seat in most districts. This implies low levels of intercoalition competition, which in turn allows coalitions to present candidates responding to strategic national interests instead of voters’ preferences. Also, the congressional overrepresentation of the right-wing coalition, together with the existence of institutional senators, protected the status quo left behind by the Pinochet dictatorship for 16 years, until the constitutional reform of 2005 did away with the so-called undemocratic enclaves. 16

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

in the first ballot. This system induces the nomination of moderate candidates who seek the median voter position. In the context of only two large coalitions having congressional representation, this system induces intracoalition negotiations to nominate one candidate for each coalition, which in turn reinforces coalition cohesion and discipline.18 Municipal elections in Chile are held nationwide on the same day. This makes them another arena for the electoral contest between the coalitions, and provides another opportunity for intracoalition negotiation to agree upon a common list of candidates (which distributes seats for each party.) The combined effect of the three electoral systems on politicians, parties, and coalitions is a set of incentives that cross-pressure them and multiply the opportunities for electoral competition and intracoalition negotiation. There is a twelve-year electoral cycle for concurrence of presidential and legislative elections; municipal elections are never concurrent. Hence between 1989 and 2005, there have been ten different electoral contests, no more than three years apart, and often held every other year or even in consecutive years, as in 1992–93, 1996–97, 1999–2000–2001, and 2004–2005.19 Different electoral systems functioning under a continuous electoral cycle have a moderating effect over the coalitions. The binomial system is the core of the electoral system, and is the main force leading parties to coalesce. However, the existence of municipal and presidential elections mediates the moderate polarizing effect of the binomial system, which induces parties to design their electoral platforms aiming at one-third of the electorate, and encourages coalitions to aim closer to the median voter. On the other hand, the high sequence of elections forces parties

In the 2006 elections, however, the Alianza presented two candidates, using the first ballot as a de facto primary. In the second ballot, both parties of the Alianza supported the same candidate. 19 The constitutional reform of 2005 reduced the length of the presidency to four years and introduced simultaneity in congressional and presidential elections. This will facilitate the operation of coalitions, allowing for “compensations” in congressional elections to the parties that must resign their presidential options within each coalition. 18

177

178

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

to successfully reach intracoalitions agreements, reinforcing coalition discipline and increasing the power of party leadership, and hence party discipline as well. Finally, the informal rules that determine the actual workings of congress reflect the power of both parties and coalitions. The two main institutions that shape the workings of congress are the executive boards (mesas) of each chamber and the chambers’ committees. Formal rules determine that committee assignments are nominated by the mesas and ratified by the respective chambers, and shall proportionally reflect the partisan composition of each chamber. In actuality, however, the mesas consider the parties’ requests for committee assignments. Proportionality is not strictly maintained, but coalitions work out agreements. Also, the chairmanship of committees is not proportionally assigned, but is agreed upon by the coalitions and normally rotates from one congress to the next. In this respect, it can be argued that the workings of congress reinforce the importance of both parties and coalitions, and that both coalitions and parties as well are central to Chilean politics—contrary to Carey (2002a), who states that the only relevant unit of analysis is coalitions.20 An Agenda-Setting Executive with an Array of Negotiating Tools Confronted by Several Checks and Balances

The Chilean presidential system has one of the strongest agenda-setting presidents in Latin America. The executive has exclusive legislative initiative on several policy areas, has a highly hierarchical control of the budget process, and has an array of urgency and veto options, which makes it a de facto agenda setter. Yet there are a number of institutional actors that are able to block executive policy initiatives (Londregan 2000). These include an independent judiciary, a constitutional tribunal, a comptroller general, and until the constitutional reform of 2005, the “National Security Council” (COSENA), which gave the armed forces For evidence of the interplay between coalitions and political parties in Chile, see Aninat and Londregan (2004). 20

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

a direct institutional role in the government. The bicameral congress, which until 2005 had almost a dozen nonelected senators in the upper chamber, requires supermajority thresholds to change many laws, enabling various minorities to block policy changes. Until 2005, presidents were elected for six-year terms and were prevented from running for immediate reelection. During the past 16 years, the process of nominating candidates has varied across elections and parties, but the two main coalitions have presented candidates for every presidential election. Despite the fact that in all elections several candidates have competed, in all three elections the candidates of the main two coalitions—which include all the parties with congressional representation—have dominated the elections. The four presidential elections have been won by Concertación candidates. The constitutionally mandated agenda-setting powers enjoyed by the Chilean president are quite substantial. Issues are divided into “matters of law” and “matters of administration.” All legislation related to the daily running of the government is considered a “matter of administration.” Members of congress can only propose legislation that is considered a “matter of law.” This restricts the legislative initiative of congress. Within matters of law, the executive has the sole legislative initiative over legislation concerning the political and administrative divisions of the state, its financial administration, the budget process, and the selling of state assets. The executive also has sole initiative in areas such as taxation, labor regulation, social security, and legislation related to the armed forces. Therefore, the executive has sole authority to initiate legislation that requires budget increases or allocation of new funds, which gives it exclusive legislative initiative over most economic policy areas. Furthermore, until 2005, the executive could convene an extraordinary legislative period, during which the executive determined the legislation congress was allowed to consider—in addition to the constitutionally mandated legislative period from May 21 to September 18 of each year.21

In the absence of a presidential request for an extraordinary legislative period, congress could convene for an extraordinary legislative period, in which case the legislation to be considered was determined by the legislature. 21

179

180

Crist贸bal Aninat, John Londregan, Patricio Navia,

and

Joaqu铆n Vial

Perhaps the constitutional provisions that best reflect the strong powers vested in the presidency are those governing the budget process.22 The formal rules governing the budget process give strong powers to the executive, making Chile one of the countries with the most hierarchical budget institutions in Latin America (Alesina and others 1999; Vial 2001). The constitutional responsibility for the financial administration of the state belongs to the president via the finance minister, assisted by the budget director. They are in charge of setting the spending limits and leading the negotiations with the spending agencies during the preparation stage of the annual budget process, and of overseeing the execution and control stages. Legislators are not allowed to introduce amendments that raise spending or create any financial commitments. Every legislative proposal by congress that has financial implications must be accompanied by a financial report prepared by the budget office (ministry of finance) with an estimate of the financial impact of the project and the sources of financing under the current budget law. The Public Finances Committee of each chamber must review the specific articles carrying financial implications. The executive has the sole responsibility for the overall revenue estimates and the presentation of a medium-term macroeconomic program, which is presented to the Special Budget Committee of congress by the finance minister at the opening of the budget debate. Congress cannot increase expenses in any item; it can only reduce or reject expenses proposed by the executive, as long as it does not interfere with the ability of the government to run state policies or meet previous legal commitments. Additionally, congress has 60 days to approve the budget law sent by the executive, or the original proposal becomes law. The tight control of the executive over the budget process, and all matters of legislation with fiscal impact, gives it control over side payments to compensate members of congress negatively affected by new legislation. For example, the executive was able to successfully implement a major

For a detailed discussion of the political economy of the budget process in Chile, see Aninat and Vial (2005). 22

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

port reform and to sign a partial accession to Mercosur, given its ability to financially compensate the losers of those reforms.23 In addition to its agenda-setting powers, the executive controls the flow of legislation through the use of “urgencies”—a constitutional mechanism designed to give the executive the power to force the legislature to vote on an initiative within a fixed time limit. There are three different types of urgencies that can be introduced and withdrawn solely by the executive at any moment during the legislative process: simple urgency, summa urgency, and immediate discussion. They require the chamber in which the bill is being considered to vote the bill in thirty days, ten days, and three days, respectively. If the bill is already in a mixed senate-chamber of deputies committee, the indicated timeframes must be divided between the vote in the mixed committee, and the subsequent votes in each chamber of congress. Therefore under simple urgency, the mixed committee and each chamber have ten days to vote the bill; under summa urgency, the mixed committee has four days to vote the bill, and each chamber has three days to do it; and under immediate discussion, the mixed committee, the senate, and the chamber of deputies have one day each to vote the legislative initiative. The executive can also exercise a wide range of veto options.24 If a president vetoes a bill, or part of it, congress needs a two-thirds majority of members present in both chambers to insist on its preferred version of the bill. In addition, the use of the amendatory veto (veto sustitutivo) gives the president the power to amend legislation that has already been passed, thus allowing the executive to exercise a last-minute bargaining tool with the legislature after the preferences of both chambers have already been made public in the floor vote of the legislation. For the approval of the partially amended vetoed bill, however, the president needs the favorable vote of a simple majority of both chambers of congress. If a majority is not reached, the amendatory veto is equivalent

For more on this, see the working paper version of this chapter (Aninat and others 2006) and Aninat and Londregan (2004). 24 See Ley Orgánica Constitucional del Congreso Nacional, http://www.senado.cl/ site/institucion/normativa/ley/. 23

181

182

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

to a simple partial veto, and the bill is approved without the contested articles. As a consequence, this power has seldom been exercised in the last 15 years, except to amend technical errors in the legal text (Aleman 2003). The president’s nonlegislative powers include broad authority to nominate, appoint, and dismiss government officials. The president nominates the supreme court justices, the central bank’s governing board members, one of the seven members of the constitutional tribunal, and the comptroller general, who must be ratified by a supermajority of votes in the senate. This procedure ensures nonpartisan—or at least politically balanced—nominations to these key institutional posts. The executive also had the authority to nominate two of nine so-called “institutional senators” every eight years, and all former presidents who serve for at least six years were entitled to a lifetime voting seat in the senate25 (more on this below, in the section on the undemocratic enclaves of the 1980 Constitution). The president directly appoints all cabinet ministers, regional and provincial governors, ambassadors, and heads of government agencies and state companies. Given that the party system fosters coalition governments, cabinet formation might be negatively affected by reducing the cabinet’s cohesion. However, the highly hierarchical structure of the cabinet, combined with the president’s authority to appoint and dismiss the cabinet ministers and the prominent role of the finance minister in determining the budget of each ministry, has produced highly efficient cabinets in the four administrations since the restoration of democracy. There has been a stable number of ministries and a low rotation of ministers and undersecretaries (there were eighteen ministries until 2004; since then, three new ministries have been created). The president, however, could not exercise discretionary power over the high command of the armed forces until the constitutional reform of 2005. The president could not remove the heads of the armed forces without the approval of the COSENA. The heads of the army, air force, navy, and Carabineros (national police) are appointed by the president Former President Aylwin did not have that capacity, since he held office during a special transitional term of only four years. 25

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

for four-year terms from a list consisting of the five highest-ranking officers of each branch. The degree of support for the Chilean president in congress is relatively high and constant under the current electoral system. As mentioned above, the binomial electoral system encourages coalition formation, and leads to rather similar congressional representation for each coalition in both chambers of congress. Therefore, support for the executive in congress oscillates around 50 percent in all legislatures. Because presidents are the de facto leaders of the multiparty coalitions that have characterized Chilean politics since 1989, presidents exercise significant influence over the decisions made by the government coalition (Carey 2002a; Montes, Mainwaring, and Ortega 2000). In all legislative and municipal elections, sitting presidents have been called upon as the ultimate arbiter in disputes among party leaders over the composition of the Concertaci贸n electoral lists (Carey and Siavelis 2003). Because they assume the role of coalition leaders, presidents have actively sought to distance themselves from the political parties they belong to in order to avoid the impression that they are benefiting their own parties and hurting the other parties that comprise the government coalition. Offsetting the powerful Chilean president is the national congress. Although it is weak in comparison to the U.S. Congress, the Chilean legislature is unusually professional and technically competent by Latin American standards. Embedded in a well-designed system of checks and balances, the legislature has an active and decisive role in the Chilean policymaking process. The bicameral legislature consists of a 120-seat chamber of deputies and a senate with 38 members (in addition to the 9 institutional senators). Deputies are elected to four-year terms, while senators serve for eight-year terms, both elected in two-member districts, with no term limit provisions. The Chilean legislature makes extensive use of committees. Only the Finance Committee is required by law to exist; other committees are established by each chamber at its own discretion, though traditionally there are 19 permanent committees in each chamber. The legislative committee system in use in Chile gives committees less power to influ-

183

184

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

ence the final composition of a bill than the U.S. House of Representatives sometimes grants its committees under closed rules provisions. Moreover, senate rules permit the floor to easily overturn committee decisions (the signatures of ten senators suffice to force a floor vote on a provision rejected in committee). The legislative process consists of three “constitutional” steps (trámite constitucional). The first constitutional step begins when a legislative initiative is submitted by the president to a chamber of his or her choice. Individual legislators and groups of legislators can also submit legislative initiatives to their respective chamber. The chamber assigns the initiative to one of its established committees or to a specially assigned joint committee depending on the nature of the initiative. The committee then discusses the bill “in general” and can modify the initiative at will. Regardless of the committee’s vote on the initiative, the chamber votes on the amended initiative presented to the floor by the committee. If amendments are offered during the general discussion, this triggers a second reading of the bill, called discussion in particular. At this stage, the bill is returned to the committee, and all amendments are discussed. Typically the chamber will agree to extend the deadline for presenting amendments. Once this has passed, the committee meets, discusses, and votes on the new amendments. The bill, along with any adopted amendments, then returns to the floor. In the senate there are easy procedures to “renew” amendments rejected by the committee (with the signatures of ten senators, less than a quarter of the chamber). Any member of the legislature can introduce amendments in advance of the second reading, and in the senate, committee members must put each proposed amendment to a publicly reported vote. Once the first legislative step is completed, the initiative moves on to the other chamber for the second legislative step. A similar process ensues. After this stage, if both chambers have approved exactly the same version of the bill, it is sent to the president for signature or veto. However, if the chambers have approved distinct versions of the bill (if the revising chamber has sustained any amendments, for example), the bill is returned to the chamber of origin, which votes on the bill article by article. If differences persist, the bill is referred to a conference com-

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

mittee made up of equal numbers from both chambers and chaired by a member of the senate. The senate conferees are typically the members of the relevant committee. Deputiesâ&#x20AC;&#x2122; committees tend to be larger (thirteen members, compared to five for the senate), so not all the deputies from the relevant committee will be involved in the conference committee. The bill put forth by the conference committee is put to an up or down vote in each chamber, without the possibility of amendment. If and when both chambers have approved the same version of the bill, it goes to the president, who can sign the bill, in which case it often goes to the constitutional tribunal for approval of its consistency with existing constitutional provisions, if a minimum number of members of congress or the president ask for it, and if provisions about the constitutionality of the objected part of the bill have been raised during the legislative process. (Objections at this stage tend to be genuinely related to constitutional issues, so that the constitutional tribunal does not act as a third legislative chamberâ&#x20AC;&#x201D;as it could if it were to abuse its authority.) As mentioned, if the president does not wish to sign the bill, he can veto it in whole, in part, or offer amendments and send it back to congress. If the veto is overridden, or if the amendments are accepted, the bill is promulgated. If a successful presidential veto applies only to part of a bill, then the remainder of the bill is promulgated. Constitutionally mandated supermajority thresholds for some special legislation contribute to offsetting the executiveâ&#x20AC;&#x2122;s substantial agenda-setting power. They range from an absolute majority of the total membership (as opposed to a majority of those present) up to two-thirds of the members.26 Constitutional reforms and laws that interpret the constitution require the highest supermajorities. A high threshold of four-sevenths of both chambers is required to reform organic constitutional laws, which

There are four types of laws in Chile. A simple majority of the members of each chamber of congress present at the time of the vote is required for ordinary laws. Laws of qualified quorum require a majority of the total number of members of each chamber. Organic constitutional laws require a supermajority of four-sevenths of the total number of members of congress. Laws that interpret the constitution and constitutional reforms require a three-fifths supermajority; to reform some chapters of the constitution, a two-thirds supermajority is required. 26

185

186

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

cover a wide range of issues, such as education, the electoral system, regulation of political parties, and the laws that regulate the central bank and the comptroller general. Social security issues also require a special quorum of an absolute majority of the total membership of both chambers. At the core of the policymaking process in Chile since 1990 has been the combination of the binominal system—which induces the formation of party coalitions that regularly win between 40 and 60 percent of the seats in congress—with the supermajority provisions. Those provisions permit the minority coalition to block legislation, forcing the government to build consensus across coalition lines in order to approve its legislative agenda. With the exception of ordinary legislation that requires the weakest thresholds, the Concertación has always faced the need to negotiate changes to legislation that require supermajority approval with the conservative opposition in at least one legislative chamber. In sum, the Chilean executive has strong agenda-setting powers and good negotiating tools that allow it to implement most of its preferred legislation. But the Chilean president cannot bypass congress through its veto, decree, or plebiscite powers, as it is the case in other Latin American countries (Aninat 2007). Besides the budget bills, which have special legislative procedures, all legislation needs at least a majority of both chambers to be approved. Independent and Politically Insulated Enforcement Technologies and Other Veto Players

In addition to the legislature, the Chilean political system has several constitutionally mandated checks and balances offsetting the executive’s substantial powers. Of the number of veto players the executive faces, some are “traditional” checks and balances present in most of today’s successful democracies, and others were “undemocratic enclaves” entrenched by the Pinochet regime in the 1980 Constitution. They were in place from 1990 until the constitutional reform of 2005. Among the former are the bicameral congress with supermajority provisions for some legislative matters that were analyzed in the previous section, and several enforcement technologies such as the judiciary, the constitutional tribunal, and the

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

comptroller general. These institutions are independent and politically insulated. They do not play any active role in legislation, but are strong independent checks on the executive. Among the latter were the semiautonomous armed forces; the National Security Council (COSENA), in which the armed forces were heavily represented; and the institutional senators (nine nonelected senators appointed by different state institutions, which gave undue political power to the armed forces). As mentioned, the three main enforcement institutions of the system are the judiciary, the constitutional tribunal, and the comptroller general. The 1980 Constitution establishes their absolute independence from other powers of the state, and their institutional design successfully accomplishes that goal. Even though all three institutions are nominated by other branches of the state, the correct combination of checks and balances assures their political independence and avoids partisan biases in their composition. In all cases, the people who aspire to hold these offices cannot cater to the interests of one branch of the state alone and, once nominated, the stipulation that they cannot be removed from their posts insulates them politically so they may carry out their duties. The judiciary, headed by the supreme court, is the final and most important enforcement institution. It has to review all legal disputes in the country, enforce property rights, and hear all cases brought by the comptroller general regarding the legality of government actions. Even though the judiciary in Chile does not have judicial review powersâ&#x20AC;&#x201D;and hence has no legislative roleâ&#x20AC;&#x201D;as the main enforcement technology, it is an important check on the executive for policy implementation. The judicial branch is composed of the supreme court, the courts of appeals, and ordinary courts. The supreme court is the highest tribunal in the country and is composed of 21 judges. They are nominated by the president from a five-person list proposed by the supreme court, and must be approved by two-thirds of the senate. Judges cannot be removed until they are 75 years old, unless sanctioned for misdemeanors. There are 17 courts of appeals throughout the country, and their judges are designated by the president from a three-person list proposed by the supreme court. The constitutional reform of 2005 increased the composition and strengthened the powers of the constitutional tribunal (TC). Before the

187

188

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

reform, the TC consisted of seven members—three supreme court judges and four lawyers—who were appointed in staggered terms, served eightyear terms, and could not be removed unless impeached. A quorum of at least five members was needed to convene. The supreme court nominated the three judges through simple majority voting in successive and secret ballots, the president and the senate nominated one lawyer each, and the COSENA nominated two lawyers. Following the reform, the president and the supreme court now nominate three members each and the senate nominates four members. These nominating procedures ensure the tribunal’s technical capability and avoid partisanship in its rulings. The rulings of the tribunal are not subject to appeals of any kind, though it can rectify previous rulings if it independently decides to do so. The Constitution of 1980 mandated that constitutional review for preventive control be performed by the constitutional tribunal, and for repressive control (ex post review), by the constitutional tribunal together with the supreme court. In practice, however, given that the Chilean legal system is based on civil law, supreme court rulings do not set a valid precedent for other similar cases, giving the constitutional tribunal de facto sole power to review for ex ante constitutionality of laws. The constitutional reform established that both ex ante and ex post constitutional reviews be performed by the constitutional tribunal.27 The TC must review the constitutionality of all organic constitutional laws and laws interpreting the constitution before they are promulgated. As mentioned, it can also review specific articles of any bill if requested by a certain number of members of congress during the discussion of a bill. In that way, the constitutional tribunal exerts ex ante jurisdiction over legislation, so it can prevent legislation from being enacted if it violates the constitution. For example, in 2001 the TC declared unconstitutional a bill to reform the pension system that was a high priority of the Lagos administration and that had been under legislative discussion for six years. The government had to redraft the bill almost entirely to overcome that ruling and carry out the pension reform (Aninat 2006). Constitución Política de la Republica (CPR), article 93, numbers 1, 3, 4, 6, 7, and 16. 27

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

Given its nonpartisan composition, the tribunal does not take an active role in shaping legislation by itself. However, it can be used by legislators to delay the discussion of a bill. A small number of legislators can temporarily stop the discussion of the bill by sending it to the tribunal to review its constitutionality, as was the case in the pension reform. The Contraloría General de la República consists of the comptroller general and the Contralorías regionales. The comptroller general is appointed by the president with the approval of three-fifths of the senate, and cannot be removed until she or he is 75 years old, unless impeached by the senate based on a constitutional accusation by the chamber of deputies. The comptroller general controls the lawfulness of the state administration. It oversees the centralized and decentralized organs and services of the state, as well as any private entities that deal with the state and receive public funds of any kind. The comptroller general verifies the constitutionality and legality of the actions of administration, but it does not have punishing powers; it proposes measures to the appropriate authorities or presents a case to the courts. It regularly publishes bulletins setting the correct interpretation of norms and administrative procedures. The comptroller general is in charge of the general accounting of the nation and calculates the annual balance of the financial administration of the state. It also provides information and advice to the executive and congress. Several nondemocratic veto players exerted a significant role in the Chilean policymaking process until their removal by the constitutional reform of 2005. The armed forces were given a political role as “guarantors of the constitution”—whatever that may mean—by the 1980 Constitution. Among the undemocratic means through which the armed forces could interfere with the executive’s policy agenda were their semiautonomous status (the president could not dismiss the head of the four branches of the armed forces at will, and they could dispose of their budgets without government oversight),28 their role in the COSENA, and their power to nominate four of the nine institutional senators.

Even though the constitutional reform eliminated the political role of the armed forces, they are still entitled to receive 10 percent of the revenues of the state-owned copper company, CODELCO. 28

189

190

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

The National Security Council (COSENA) was also drastically modified by the constitutional reform, leaving it as a consulting body for the president. During the previous 16 years, however, it wielded significant influence in Chilean politics. It was comprised of eight members: the president of the republic, the president of the senate, the president of the supreme court, the comptroller general, and the heads of the armed forces. The COSENA appointed four of the nine institutional senators (who have to be themselves former heads of each branch of the armed forces) for eight-year terms to the highly contested and narrowly divided senate, and also appointed two members of the constitutional tribunal. While the COSENA did not take an active role in the formulation of economic policy, it was actively engaged in shaping the policies related to the human rights record of the former military regime, including the government’s response to the arrest of General Pinochet in the United Kingdom in 1998. The nine “institutional senators” were selected as follows: two were chosen by the outgoing president, three by the supreme court (two had to be former supreme court justices, and one a former comptroller general), and four were appointed by the COSENA. The outgoing dictatorship nominated all nine institutional senators in 1990, giving the rightwing coalition a majority in the upper chamber throughout the 1990s and early 2000s without commanding a majority of the popular vote. As time passed, the Concertación administrations were able to nominate some institutional senators. In 1997, President Frei nominated two senators, significantly altering the composition of the senate for the remaining two years of his term, and in 2000 he became senator for life. The constitutional reform of 2005 eliminated the institutional senators and allowed the Concertación to have a legislative majority in both chambers of congress for the first time since the return to democracy. In sum, the combined effect of a de facto agenda-setting president, who has effective negotiating tools but is confronted with several checks and balances, produces a policymaking process in which legislation is not easily approved. But the very difficulty of approving legislation means that once a measure is approved, it is hard to overturn. Passing a law in Chile represents a genuine policy commitment.

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

Capabilities of the State Bureaucracy

Chile is a unitary state; its main political divisions are 13 regions subdivided into provinces. The administration of the state is centralized in the central government in Santiago, with local branches of the central government located throughout the country. There is some degree of local decision making at the municipal level, but there is no fiscal decentralization. Communities have some opportunity to participate in local decision making through consulting councils at the regional and provincial levels. The head of the administration of the state is the president of the republic. The bureaucracy is divided into ministries for sectoral administration, and intendances (Intendentes) and governorships for regional and provincial matters, respectively. The president can appoint and dismiss at will all ministers, all Intendentes, and all governors. Coordinating sectoral decision at the regional and local levels are the SecretarĂas Regionales Ministeriales (SEREMIs), which administratively depend on the ministries but work with the Intendentes to implement the sectoral policies at the regional level. Finally, there are councils for local consulting at different levels, which incorporate local concerns, but have no decision-making power. Chileâ&#x20AC;&#x2122;s bureaucracy has undergone a profound process of reform in the last 30 years. During the twentieth century it developed a reputation for low levels of corruption by Latin American standards, but also was marked by strong centralization, an emphasis more on procedures than on outputs, and no participation of civil society or market mechanisms in the provision of public services, making it rigid and not very efficient. The major economic transformations carried out by the military dictatorship in the mid-1970s started a slow but incremental process of reform that led to major reforms in the late 1990s and early and mid-2000s during the Frei, Lagos, and Bachelet administrations. The military implemented structural economic reforms and reduced the size of the state by privatizing public companies and the social security system, deregulating several markets, and improving the tax system. These reforms, however, were carried out in an ideological environment contrary to empowering

191

192

Crist贸bal Aninat, John Londregan, Patricio Navia,

and

Joaqu铆n Vial

the state, which led to underinvesting in state capabilities and no effort to modernize the provision of public services. Since the end of the dictatorship, as both the nascent democracy and the new economic institutions have gained popular support and consolidated, a gradual reform of the civil service has been taking place. The Frei administration initiated a significant modernization of the state, which has been further developed by the Lagos and Bachelet administrations after corruption scandals broke out in 2003 and 2006. The most important pieces of state reform legislation approved during the Lagos administration addressed the following issues: simplifying administrative procedures (Law 19.880), updating the level of remuneration of high public officials to make it more in line with the private sector, reducing the amount of undisclosed funds (Law 19.863), professionalizing the public service in order to improve and develop long-term careers in the civil service regardless of political changes (Law 19.882), and increasing the transparency of government procurement and the concessions system (Law 19.886). In November 2006 the Bachelet administration introduced to congress a wide-ranging legislative package focusing on anticorruption and transparency. Its main focus was on regulating lobbying activities, political campaign finance, and conflicts of interest among members of congress; legislating to protect the right to access to public information; deepening the civil service reform; and modernizing the comptroller general. As of July 2007, these bills were under legislative discussion. Other bills still under discussion in congress are intended to decentralize the execution of social spending, modernize municipal financing, reorganize the state holding of public companies, and improve the capabilities of regional governments. Changes to the administration of the state that do not require legislation and that are already in place include a wide range of e-government initiatives,29 and improvements in the performance indexes for civil servants. The latter were introduced in the mid-1990s as experimental prowww.gobiernodechile.cl, www.presidencia.cl, www.elecciones.gov.cl, www.chilecompra. cl, www.dipres.cl, www.senado.cl, www.camara.cl. 29

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

grams, but have evolved to cover all central government institutions and are currently used to partially link salaries to measures of performance. They are regularly scrutinized by congress during the annual discussion of appropriations as part of the budget allocation process. Despite the ongoing effort to improve the capabilities of the state, Chile’s bureaucracy is still rigid and procedure-oriented, and wages at the professional and high-responsibility levels in the public sector are lower than their counterparts in the private sector. These shortcomings lead to lower performance and at the end of the day contribute to a lower quality of public policy. It is important to emphasize that these criticisms of Chile’s civil service are based on a comparison with OECD best practices and that the Chilean bureaucracy is highly competent by Latin American standards. Interactions: The Policymaking Process in Chile

While numerous features of the Chilean system have thus far been mentioned, two are most salient from the standpoint of implementing cooperative policies. The first is the existence of a small number of long-lived political parties that interact repeatedly with one another and with the voters. The second is the existence of implementing institutions—the bureaucracy and the judiciary—that function honestly and with some degree of efficiency and transparency. Repeated interaction between the parties not only makes it possible for them to make (and keep) policy deals, but more importantly it creates an incentive for the parties to maintain their ideological “brand names” with the voters—thus constraining the sorts of policy changes with which they align themselves. Because the laws passed and executive decrees handed down are actually implemented, deals are credible, opening the door for policy agreements that involve intertemporal trades. While other noteworthy institutional features such as the agenda-setting powers of the executive, the number and disposition of the veto players, and the degree to which policymakers’ interests align on a given issue are important to the details of the policymaking process, Chile would achieve a much less cooperative policymaking process without

193

194

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

its party structure and implementation technology. Neither changes in the executive’s agenda-setting power, nor in the blocking capacity of the various veto players (insofar as it did not interfere with the equilibrium party structure or the integrity of the judiciary or the bureaucracy), nor a variation in the degree of interest-group alignment would rule out at least some degree of cooperation in policymaking. To illustrate the relative impacts of the salient features of the Chilean PMP, it is useful to consider some examples of policy compromises. Two important reforms—the partial privatization of Chile’s major port facilities (Law 19.542) and the approval of Chile’s special relationship with Mercosur—each led to improvements for the general public, and each came at the cost of a disadvantaged group. In neither case did the disadvantaged group enjoy a policy veto, yet in both cases the government committed itself to compensating the losers from the reform. In the case of port modernization, the government faced the need for an expensive upgrading of Chile’s outdated port facilities. The massive capital outlay required would have placed enormous strain on the public budget, and so the government chose instead to grant long-term concessions to private investors in return for the investors undertaking specific construction projects. This route actually relieved the strain on the budget, since the private concession holders paid for the privilege of running the port facilities. The government calculated that the higher rates charged by the private service providers would be compensated by a more rapid processing of ships through the port, leaving overall shipping costs lower. This has happened; transfers from port operators to the government have far exceeded initial estimations. However, this reform entailed replacing the existing state-owned enterprise EMPORCHI, which controlled most of Chile’s ports, and it meant that existing EMPORCHI employees would lose their jobs. Given that the great majority of Chileans would be better off as a result of the proposed reform, the government could have simply scrapped EMPORCHI, lived with the ensuing strike, and gone on with the reform. However, the dominant Concertación coalition of left-of-center parties could only choose to abandon the EMPORCHI workers at the cost of vio-

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

lating some of their basic ideological premises regarding the treatment of workers, thus sending a signal to the rest of the labor movement that the government was not to be trusted. Because the Concertación parties were in a long-term relationship with organized labor, which they needed for electoral purposes, it was politically costly to simply abandon the port workers. Accordingly, the port reform included expensive “parachutes” for the dismissed dockworkers, such as educational benefits for their children, occupational retraining, and funds for dismissed longshoremen to start small businesses. The right-of-center opposition was much less ideologically constrained to compensate the dock workers, and the president could have formed a coalition of opposition legislators and some members of his own party coalition to push through the reform without expensive concessions to the dock workers. However, the constraining effects of ideology and of party reputation prevented the president from doing so. A similar process played out in 1996 on an even larger scale with the passage of the government’s trade agreement with Mercosur. This treaty would open attractive markets for Chilean pharmaceutical manufacturers and reduce the price of food in Chile. The Mercosur countries were among the world’s most efficient producers of vegetable oils, wheat, and sugar. This threatened Chile’s “traditional” agricultural sector, which was involved in the production of sugar beets and wheat. Both the Concertación and the opposition Alianza were electorally competitive in the adversely affected regions, so the government risked long-term alienation among adversely affected farm voters and eventual seat losses if it pushed the policy through with a majority coalition of legislators from outside the affected areas. Instead, the government persuaded almost all of its adversely affected legislators to vote in favor of the treaty, promising economic compensation to the affected farm areas.30 This compensation actually arrived in the form of a budget reallocation equal to 3 percent of the 1999 budget, and in the form of redesigned price supports in 2003 that had the effect of extending temporary tariff protection to sugar beet The fact that electoral districts in the affected regions were heavily biased in favor of rural areas was also a key factor. 30

195

196

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

growers. These latter shifts of resources were not trivial (the sugar tariff fell disproportionately on the poor) but the Concertación sustained them in order to preserve its reputation; long-term interaction with the voters made reneging too costly. In the case of the legislative approval of the government budget, the executive enjoys stronger powers relative to congress than in other policy areas. Nevertheless, congressional approval is required to pass a new budget (but if congress does not address the government’s budget on time the government’s initial proposal automatically becomes law). One possible outcome of Chile’s institutional structure would be for the opposition to follow a “scorched-earth” policy of voting against the government’s budget and forcing a constitutional impasse, given the practical ambiguities of the constitutional norm. However, this is not what happens; instead, budgets typically pass with majority support from both the government coalition and from the opposition. This is partly facilitated by rules limiting the government’s discretion in allocating funds, and by the independent supervision of expenditures by the Contraloría General. Any efforts by the government to act in bad faith regarding budgetary allocations would be quickly detected. But why would there be good faith to begin with? First of all, it exists because even the opposition wants to avoid a constitutional crisis caused by gridlock.31 Moreover, the actual budgets reflect compromise solutions with the opposition, in which spending growth for favored government programs is slower than the executive would prefer, giving the opposition something to point to as an achievement and restraining the growth of spending, even as the government can deliver a balanced budget and pass its programs. Without the ongoing reputations of the ruling coalitions, the temptation for individual legislators to play “blame games” by sabotaging new budgets and arguing over who was responsible might become overwhelming. Voters would then be stuck trying to sort out which maverick legislators were really to blame and which were in fact

Something that is often on the minds of members of congress when the final deadline approaches is the memory of the 1891 Civil War (the bloodiest conflict in Chilean history), which started after congress rejected President Balmaceda’s budget. 31

Political Institutions, Policymaking Processes,

and

Policy Outcomes

Chile

defending voters’ interests. The organization of the legislature and the executive within the rubric of two electoral coalitions makes allocating blame much easier, and so increases the incentives for politicians to implement cooperative solutions. Looking across policy areas, one can see that when the executive has greater control over policy, as in the budgetmaking process, policies are closer to the executive’s preferred outcome. Likewise, an alignment of interests facilitates the passage of legislation and reduces the probability of gridlock, which delayed resolution of the military government’s human rights legacy, and which has hampered reform in areas such as education. As education reform requires a four-sevenths supermajority of both houses of congress for its approval, the Bachelet administration has been unable to pass one of its top priority reforms. However, even when the president is in a position to impose his or her will, he or she often makes concessions to adversely affected minorities. These concessions arise from the governing coalition’s need to maintain its ideological “brand name” and its reputation for keeping its policy promises (as in the case of its favorable treatment of sugar beet farmers). The need to maintain a reputation, for the sake of ideological consistency and for the sake of being able to keep one’s word, stems from the stability of Chile’s political parties and their small number. This cooperative behavior is facilitated by the existence of good institutions for enforcement and implementation, and it responds, on margin, to variations in the institutional structure of the legislative process and to differences in the alignment of interest. However, it is the organization of policymaking around a small number of long-lived parties that is a decisive factor in the cooperative nature of policymaking in Chile. Conclusion

The policymaking process in Chile is characterized by an institutional structure in which a gauntlet of institutional veto players opposes an agenda-setting executive. The institutions responsible for enacting policies into law are relatively transparent, honest, and efficient, though there is significant variation on this last score. The electoral system (which is

197

198

Cristóbal Aninat, John Londregan, Patricio Navia,

and

Joaquín Vial

open-list d’Hondt with two member districts, and has selection within lists by plurality rule) has fostered the development of two ideologically distinct coalitions of political parties. The small number of parties (there are five major parties) and their longevity make them vehicles for accountability, staking out credible ideological positions with the voters and cultivating reputations for fulfilling promises. Within this framework, the relative divergence or alignment of interests by the major actors influences the ease with which cooperative outcomes are reached. The legislative institutions create pressures for selective gridlock, punctuated by policy changes that strain the tolerance of the veto players. However, the capacity of the system to actually implement the laws that are passed makes negotiations over policy outcomes meaningful: policy debts are paid in the hard currency of policies that will actually be carried out, with verifiable contents. The party system, which is shaped by the electoral laws, creates an environment in which policymakers can break out of the structure of an agenda-setting president and a gauntlet of veto players to implement cooperative policy choices. Because there are relatively few parties and because they are longer lived than the individual politicians, they are willing to invest in their long-term reputations, which facilitates intertemporal bargaining. While any change in the institutional details would influence the behavior of the system, we believe that it is the political party system that is the essential foundation for cooperation. Change other features of the process (by weakening the executive’s agenda-setting power, or by removing some of the veto players, for example) and the parties (and voters) would adapt, but they would continue to cultivate their reputations, and this would lead them to continue to act as guarantors of intertemporal bargains. Fracture the party system (by adopting a higher-magnitude electoral rule, or implementing a federal structure with powerful governors, for example) and the status quo legislative institutions and bureaucratic capacity would fail to support cooperative policy outcomes.

6 Political Institutions and Policy Outcomes in Colombia: The Effects of the 1991 Constitution Mauricio C谩rdenas, Roberto Junguito, and M贸nica Pach贸n

Introduction

In an influential volume edited by Rudiger Dornbusch and Sebastian Edwards (1991) on the macroeconomics of populism in Latin America, Miguel Urrutia (1991) argued that populist macroeconomics were absent in Colombia, contrary to the norm in the rest of South America. According to the definition used in that volume, one feature of populist economics is the presence of large fiscal deficits, reflecting the use of budget expenditures for redistributive purposes without a concurrent effort to raise tax revenues. Fiscal deficits in Colombia were relatively small (rarely exceeding 4 percent of GDP) for most of the past century and, when present, were rapidly corrected (see figure 6.1). Moreover, the public sector was relatively small by regional standards (aggregate expenditures and revenues of the consolidated public sector were around 20 percent of GDP between 1960 and 1990). However, since the early 1990s, fiscal policy outcomes have changed significantly. There has been a strong deterioration of the fiscal balance, as figure 6.1 also shows. The deficit of the central government has been close to 5 percent of GDP since the late 1990s, without a clear indication of a major correction. Aggregate public expenditures grew to 33.7 percent of GDP in 2003 from 21.2 percent in 1990, reflecting a deliber-

200

Mauricio Cárdenas, Roberto Junguito,

and

Mónica Pachón

Figure 6.1. Fiscal Balance of Central Government, Colombia, 1905–2005 4

Percentage of GDP

2 –0 –2 –4 –6 –8

1905 1909 1913 1917 1921 1925 1929 1933 1937 1941 1945 1949 1953 1957 1961 1965 1969 1973 1977 1981 1985 1989 1993 1997 2001 2005

–10

Year Source: Junguito and Rincón (2004) and Ministerio de Hacienda y Crédito Público.

ate effort to increase the size of the state and use fiscal expenditures for redistribution. Although total fiscal revenues grew to 29.7 percent of GDP in 2003 from 20.6 percent in 1990, the data are unequivocal on the greater tolerance toward fiscal deficits in recent times.1 In the words of Alberto Carrasquilla (2003, p. 23), “It has not been feasible to consolidate a political agreement that excludes fiscal disequilibrium as an option for public and private agents.” The presence of large and structural fiscal deficits contrasts with Colombia’s previous experience in terms of macroeconomic management.

Ocampo (2004) argues that “the Gaviria (1990–94) and Samper Administrations (1994–98) did not have the explicit purpose of raising the fiscal deficit, but rather to increase the size of the state in an orderly manner,” concluding that recent fiscal policy in Colombia cannot be characterized as populist macroeconomics in the sense of Dornbusch and Edwards (1991). 1

Political Institutions

and

Policy Outcomes

Colombia

Economic fluctuations have been much larger since 1990, while GPD growth has decelerated (to 2.9 percent per year during the 1990s from a 4.7 percent annual average between 1950 and 1990). However, inflation rates have fallen since 1990, reflecting the effectiveness of new political institutions that ruled out high inflation as an outcome (see figure 6.2). The purpose of this chapter is to understand the causes of the change in the quality of policies as well as their outcomes: improvement in some areas and deterioration in others. We focus on fiscal and monetary policies, but the analysis can easily be extended to other areas, such as trade policy and the regulation of public utilities. Rather than analyzing the details, we identify some common characteristics of policies and the policymaking process (PMP) that can be related to changes in the political institutions. More precisely, our goal is to understand how relevant political institutions such as the constitution map into political behavior, political behavior maps into policymaking processes, and policymaking processes map into policies and policy outcomes. Figure 6.2. Inflation in Colombia, 1950â&#x20AC;&#x201C;2006 35 30

Constitution in 1991

Percentage

25 20 15 10 5

2004

2001

1998

1992 1995

1989

1986

1983

1980

1977

1974

1971

1968

1965

1962

1959

1956

1953

1950

Year Source: Banco de la RepĂşblica and Departamento Administrativo Nacional de EstadĂstica.

201

202

Mauricio Cárdenas, Roberto Junguito,

and

Mónica Pachón

It is analytically tempting to focus on the effects of the 1991 Constitution, which changed the rules of the political game along multiple dimensions. The new constitution strengthened the checks and balances of the political system in an effort to endow political institutions with greater legitimacy after decades of limited participation and low representation. Although remaining extremely powerful even by Latin American standards—surpassed only by the Brazilian executive—the president lost some capacity as an agenda-setter relative to the previous period, while congress and the constitutional court (CC) gained relative power. As a consequence of the larger number of relevant players and the diminished presidential powers, political transaction costs increased in several policy areas, so the gains in representation came at a cost in terms of cooperation. In addition to the changes in the number of key players, as well as in the rules of the political game, the new constitution covered many specific, previously nonconstitutional, aspects of policy. This is particularly the case of fiscal policy, where key components of public expenditures—such as pensions, fiscal transfers for education and health, and public sector wages—were either “hard-wired” into the constitution or have been determined through the constitutional review process. Many observers have suggested that these embedded rigidities have been a major source of growing fiscal deficits and public debt, tax uncertainty, and inefficient government expenditures, which are frequently mentioned as the causes of macroeconomic instability and low growth. In an extreme form of rigidity, changes in fiscal policy often require constitutional reform, which by definition involves transactions that are slow and costly. As a result, fiscal policies are not adaptable to economic shocks, in contrast to Colombia’s own previous experience. In the case of monetary and exchange rate policies, the 1991 Constitution granted independence to the central bank’s board and banned direct monetary financing to the private sector. (Lending to the government is allowed but requires unanimity of all board members.) These constitutional provisions have been reinforced by the appointment of technocrats to the board by successive presidents. As a result, policies in this area have been more stable, coherent, and public regarding since

Political Institutions

and

Policy Outcomes

Colombia

1991. Policies have also been adaptable to external shocks, especially after the central bank’s success in bringing down inflation (there is some indication that flexibility was initially sacrificed to gain credibility). To develop the argument, the chapter proceeds in the following way. The second section familiarizes the reader with some historical background about Colombia. The third section describes the key political actors in Colombia: the president, congress, political parties, and the constitutional court. The idea is to show how the relative powers of the different players have evolved throughout time. The fourth section discusses the policymaking process in Colombia, with special emphasis on those aspects that changed after the 1991 Constitution. The fifth section presents the empirical exercises, using a database containing 3,428 bills discussed in congress during the last 20 years. The main point of the exercise is to document the relative reduction in presidential powers, measured by the increased difficulty of the executive in enacting laws in congress. More importantly, the evidence suggests that the fragmentation of congress (through more lists and more parties) explains the greater difficulty in passing laws. The sixth and seventh sections discuss in more detail the salient features or outer characteristics of fiscal and monetary policies, respectively, and illustrate the policymaking processes in both cases. The eighth section contains a summary of the main findings and some thoughts on the policy implications of two recent institutional changes: the possibility of presidential reelection and the reform of electoral rules. Historical Background: Toward the 1991 Constitution

In 1958, after ten years of political violence between the two traditional parties (Liberal and Conservador), and four years of a military government (1953–57), the parties agreed to share power during four presidential terms between 1958 and 1974, in what was called the Frente Nacional. In addition to alternation in the party affiliation of presidents, the agreement (initially a referendum and later a constitutional amendment) included strict “parity” between the two parties in the key policymaking arenas, such as congress, cabinet positions, courts, governors, and

203

204

Mauricio Cárdenas, Roberto Junguito,

and

Mónica Pachón

mayors. Many features of this agreement survived after its formal end in 1974. In spite of their apparent success in terms of economic policy outcomes, the political institutions and the consequent rules of the political game derived from the agreement were unsustainable. The left did not have access to a democratic channel for participating in the political process, weakening the legitimacy of the Frente Nacional and resulting in high rates of abstention. Popular support for the existing political institutions was also eroded by problems of patronage and corruption, as well as by the escalating violence and the growing political influence of drug money—in particular, the Medellin and Cali cartels—at both the regional and national levels.2 Capturing this discontent, and benefiting from the weak presence of the state in parts of the country, guerrilla groups—such as the M19, EPL, ELN, and FARC—gained momentum in urban and rural areas, along with labor and social movements. Early on, President Belisario Betancur (1982–86) proposed a political rather than military solution to the guerrilla problem. In 1984, his government signed a truce agreement with the FARC and started the negotiations with the M19, the two largest guerrilla groups at the time. These efforts failed, and the confrontation escalated to a higher level. In 1989, during the Barco administration, the M19 (and other smaller insurgent groups) finally laid down their arms and successfully entered the political process. However, the assassination of three presidential candidates in 1989, including the likely winner, Luis Carlos Galán, elicited popular demand for political reform.3 The long-debated idea of a constitutional reform gained popular support as a response to the political unrest, at

Drug trafficking activities and crime rates escalated during the 1980s: the production of cocaine increased by 672 percent between 1981 and 1990 (Rocha 1999), while the homicide rate rose from 42 to 89 per 100,000 inhabitants from 1981 to 1991. The negative economic implications of these trends have been well documented. See, for example, Levitt and Rubio (2005) and Cárdenas (2007). 3 The other two candidates represented the left: Bernardo Jaramillo (UP) and Carlos Pizarro (AD-M19). 2

Political Institutions

and

Policy Outcomes

Colombia

a time when great importance was given to the successful incorporation of guerrilla groups, especially the M19, into the political system.4 Ultimately, it was the students’ initiative through the movimiento séptima papeleta that succeeded in this effort.5 President Virgilio Barco, in recognition of the people’s will, issued Decree 927 of 1990 to authorize the introduction of a referendum onto the ballot of the May presidential elections calling for a constitutional assembly to reform the national constitution. The supreme court declared the constitutionality of the decree by admitting, at the same time, the state’s incapacity to confront the different types of violence facing the nation (Lleras and Tangarife 1996). The ballot, which became an official plebiscite, resulted in more than 5 million votes in favor of the national constituent assembly. President-elect Gaviria called for elections in December to organize the constituent assembly. The new constitution was adopted on July 4, 1991. Key Political Actors in the Policymaking Processes

The entire set of characteristics of the PMPs underwent important modifications during the late 1980s and early 1990s, including the role of key and veto players, the policy initiation process, and the effective number of parties. These changes mostly related to the end of the rules imposed by the Frente Nacional and the emergence of new political institutions. Since the drafting of the new constitution in 1991 is the most salient of these transformations, it is convenient to characterize key political actors and the corresponding PMPs before and after that year.6

Previous reform proposals for institutionalizing political parties, providing guarantees to opposition parties, and improving the electoral system had not prospered. See Hartlyn (1993). 5 This refers to a seventh extra-legal ballot in the March 1990 elections in favor of summoning a national constituent assembly. See Dugas (2001). 6 Some of the changes in the underlying political institutions preceded the 1991 Constitution, such as the direct election of majors (1986). Other factors have also had an impact on the PMP, such as market-oriented reforms introduced in the early 1990s and the greater interest of the United States in Colombia in recent years. 4

205

206

Mauricio Cárdenas, Roberto Junguito,

and

Mónica Pachón

The President

As has been the case in most Latin American presidential regimes, an important number of constitutional prerogatives make the Colombian president the main agenda setter in most policy areas. This has remained true even after the 1991 Constitution, which reduced presidential powers in a number of dimensions. Table 6.1 compares constitutional presidential powers before and after the 1991 Constitution. Proactive Powers. The 1886 Constitution established a highly centralized power structure, which was elevated to a new level with the 1968 constitutional reform promoted by President Lleras-Restrepo. In addi-

Table 6.1. Constitutional Presidential Powers, Colombia, 1958–2004

National Front and transition (1958–1980s)

After the 1991 Constitution (1991–2004)

Proactive powers

High decree powers, urgency petition, ex post judicial review, areas of exclusive introduction of legislation (from 1968), declaration of unrestricted state of siege and state of economic emergency

Restricted decree powers with ex ante judicial review, call for joint permanent committees along with urgency petition. Declaration of state of siege for periods of 90 days, for a maximum of 180 days, subject to approval by the constitutional court

Integrative powers Appointment powers in the governorships, various autonomous agencies, cabinet (highly centralized)

Appointment of the cabinet, autonomous agencies (highly decentralized)

Reactive powers

Required supermajority to override in economic bills (2/3 of the members of each house)

Required majority to override: 1/2 of the members of the house

Partisan powers

Relatively low due to electoral rules (no vote pooling across lists, no limit in the number of lists per party). Majority of 2/3 required in congress (until 1974).

Extremely low. No nomination power, electoral rules. Nonconcurrent elections: congressional elections first, majority runoff system.

Source: Authors’ compilations.

Political Institutions

and

Policy Outcomes

Colombia

tion to strengthening the capacity of the president to initiate legislation by augmenting the constitutionally delegated decree power, the 1968 reform significantly increased executives’ control over the budget. The president could exclusively introduce legislation in key areas, issue decree-laws when in crisis, and issue administrative decrees with limited judicial review. The 1991 Constitution deliberately curbed the legislative powers of the president by limiting to 90 days the declaration of either a state of internal commotion or a state of economic emergency (extendable for another 90 days if considered necessary). The constitutional court may revoke decrees issued, including the declaration of the emergency or commotion, if they are considered unconstitutional in any way. Previously, decrees issued under a state of siege became law, even after their special status had expired. Since 1991, decrees remain in force only if congress enacts them in regular sessions, while the pro tempore powers cannot be used to decree codes, statutes, organic laws, or taxes (see Archer and Shugart 1997). The presence of the constitutional court has restrained the use of the president’s special powers. Before 1991, the president needed only the signatures of all his cabinet members to use extraordinary powers to issue decree-laws in a 90-day time frame. Although the subject matter was supposedly narrow, in practice the president could freely use the emergency powers to make policy. The supreme court was the only veto gate, as it had the duty of reviewing each decree enacted, but it was not as active and independent as the current constitutional court. The judicial budget was determined by the executive, and justices to the supreme court were elected from lists submitted by the president to congress. A critical aspect was that magistrates had short terms of five years, with the possibility of reelection. In contrast, the constitutional court is elected by the senate only for longer terms (eight years without reelection), with only one-third of the lists submitted to the senate originating in the executive. One important aspect in which there has been no change is the president’s permanent control over the legislative agenda by using the discharge or urgency petition. This petition enables the president to prioritize a bill in the legislators’ agenda. Congress then has 30 days to

207

208

Mauricio Cárdenas, Roberto Junguito,

and

Mónica Pachón

debate and pass or reject the bill. To speed up the process, the president can also ask for joint sessions of house and senate committees in charge of the law, thereby limiting debates to two instead of the regular four. The executive has retained the exclusive right to introduce bills concerning the structure of the ministries, salaries of public employees, foreign exchange, budget, external trade and tariffs, and national debt, among other areas. Integrative Powers. Before 1991, the president had greater appointment powers (governors, justices, and heads of control entities, among others). The president also appointed the governor of the central bank. Reactive Powers. The president can either veto legislation, on procedural and substantive grounds, or suggest changes to the text approved by congress. However, congress can override the executive’s veto or annotations with a simple majority. Before 1991, a two-thirds majority was required to overrule a presidential veto. Electoral Rules. Finally, a majority runoff electoral rule replaced the plurality rule. The majority runoff, along with the timing of congressional elections (which take place before the first round presidential election), have changed the patterns of coalition building. After 1991, one can observe a clear pattern of post-electoral coalitions, and a greater role of legislators in the election of the president. Congress

The symmetric two-chamber Colombian congress can be classified as reactive more than proactive, given its lack of organization and resources and the substantial legislative powers of the executive (see Payne 1968). A powerful executive, combined with an extremely personalistic electoral system, implicitly delegates the national agenda to the president and the cabinet members. Nonetheless, the legislature has always been an important veto player for the president, especially in areas where the executive does not have complete jurisdiction.

Political Institutions

and

Policy Outcomes

Colombia

Figure 6.3. Number of Lists Running Per Election, Colombia, 1958â&#x20AC;&#x201C;2006 1000 900 800

No. of lists

700 600 500 400 300 200 100 2006

1998

1991

1986

1978

1970

1966

1962

1958

Year House

Senate

Source: Pizarro (2002).

The 1991 reform reduced the number of representatives from 199 to 165, and the number of senators from 113 to 102. The reform also transformed the national territories (large portions of the territory with low population density) into departamentos and established a minimum representation of two house members per departamento. Thus seven districts achieved representation in the house for the first time. Since 1991, senators have been elected in national, rather than departmental, constituencies.7 Electoral Rules. Congressional elections have become increasingly contested since 1991 (see figure 6.3). The rapid growth in the number of

The house has five members elected from special districts (two from indigenous communities, two from Afro-Colombian communities, and one from Colombians living abroad). Two senators are elected from special indigenous districts. 7

209

210

Mauricio C谩rdenas, Roberto Junguito,

and

M贸nica Pach贸n

lists competing in congressional elections has also implied that a large number of successful lists elect only one legislator. For example, in 2002 only three lists, out of more than 300, succeeded in electing more than one senator. To a large extent, this was the result of the electoral rules used for 40 years, until the 2003 political reform. With the formula of largest remainders (LR) or Hare system, seat quotas were calculated by dividing the number of votes by the number of seats. Seats were first allocated to lists that surpassed the quota. The remaining seats are then allocated to the largest remainders, generating incentives for parties to fragment into factions.8 The key point is that the largest remainder formula was applied in each district to factional lists rather than party lists (Payne, Adsera, and Boix 2003). The electoral law allowed multiple factions to present lists under the same party label. Thus, though the factional lists were closed and blocked, in effect the system was equivalent to unblocked party lists. The result was that parties increased the number of lists over time, maximizing their seat share while enhancing their decentralization and fractionalization. The 1991 reform did not change this situation, mainly because no representation thresholds were put in place to discourage small lists. The largest remainders-Hare system remained in place, encouraging fragmentation and election by largest remainders. Furthermore, state campaign funding is given directly to candidates, creating additional incentives to form factions or movements, instead of depending on national party resources. The combination of the incentives created by the electoral system and by campaign funding resulted in a large number of movements that were essentially electoral machines. To illustrate the point, take the case of a district with 1,000 voters and ten seats, so the quota is 100. Party A has 650 votes, party B has 240, party C has 70, and party D has 40. Accordingly, party A would get seven seats (six by quota, one by remainder); B would get two (by quota); C would get one (by remainder); and D would get none. Now, if party B subdivides into three factions, it would get three seats (all by remainder), taking one away from party A, which would get six. Party C would not benefit from fragmenting, however. This shows that this strategy is beneficial only for parties that are large enough to get at least one seat allocated by quota (in fact, if party C splits into two equal parts, it gets nothing). Now, if party A splits into eight factions and party B into three, they will capture all ten seats (eight and two, respectively). 8

Political Institutions

and

Policy Outcomes

Colombia

For example, 72 movements obtained at least one seat in the house of representatives in the 2002 election. Although these movements are typically affiliated with a major party, the excessive fragmentation of legislators was a key characteristic of congress, with an important impact on the policymaking process. The 2003 electoral reform limited the number of parties by requiring a minimum share of the total votes in order to have legislative representation (2 percent in the case of the senate). As a result, only 14 parties won representation in the congress elected in 2006. Organization. Each chamber is organized into seven Comisiones (eight before the 1991 Constitution), which have a significant role in the workings of congress. The small number of standing committees contrasts with the Mexican or Argentinean legislatures, which have more than 35 committees with overlapping jurisdictions. Legislators can belong to only one permanent committee and must remain on that committee for their entire four-year term; committee membership is determined by elections. Party membership is indispensable in order to access the committee of first preference (Pach贸n Buitrago 2003). This is important because it shows that behind a veil of fragmentation and atomization, party structures play a role in organizing legislative activity. Plenary sessions rely heavily on what is approved by the committees, especially in economic and budget matters. Membership in the economic and budget committees is highly valued because of proximity to government funds. The committee in charge of constitutional and political affairs is also highly visible, with highly valued membership. When compared to other Latin American countries, Colombia has one of the highest incumbency rates. Certain committees (such as the constitutional, economic, and budgetary committees) offer legislators an incumbency advantage, as well as a higher level of professionalization in comparison to other committees. As a result, members of those committees are the natural leaders of congress. Incumbency rates for members of these committees can be as high as 69 percent, while the rates are on average between 30 and 41 for other legislators. Professionalization thus seems to provide an advantage to incumbents.

211

212

Mauricio Cárdenas, Roberto Junguito,

and

Mónica Pachón

Political Parties

The Colombian party system is considered one of the oldest and most institutionalized in Latin America (Mainwaring and Scully 1995b). The system is characterized by high intraparty competition, functional party coalitions in congress, and significant party discipline in presidential elections. Although intraparty competition was always part of Colombia’s political history, lack of vote pooling since 1974, political decentralization since the early 1980s, and the 1991 constitutional reform brought it to new levels.9 To make politics even more candidate-based, the 1991 Constitution confirmed the independent electoral authority (Consejo Nacional Electoral), in charge of elaborating and distributing the ballot (tarjetón) with the name and picture of all aspiring candidates. Before 1986, candidates were responsible for handing out their lists on pieces of paper (called papeletas) that were used to cast a vote, and party infrastructure was useful for their distribution among voters. By 1998, 67 parties and movements were recognized by the electoral authorities and were receiving public funding for campaign expenditures (Posada-Carbó 2001). However, the internal rules of congress enhance the role of the parties and recentralize power in a few hands. Parties are determinant for accession to membership on specific committees, hierarchical positions (presidents and vice presidents of the senate, house, and their respective committees), and strategic roles such as sponsorship of key bills. Sponsors play a key role as political brokers between the executive and other legislators. Finally, congress has become more fragmented as measured by the Laakso-Taagepera index of Effective Number of Parties (ENP) shown in figure 6.4.10 The ENP increased from two political parties in the 1970s (Liberal and Conservative), to three in the post-1991 period (Liberals,

See Pizarro (1995); Archer and Shugart (1997); Nielson and Shugart (1999); Gutiérrez (1999, 2001); Rodríguez-Raga (1999, 2001); Shugart and others (2006). 10 We consolidate movements and factions into parties based on the party affiliation of their leaders. 9

Political Institutions

and

Policy Outcomes

Colombia

Figure 6.4. Effective Number of Political Parties in the House of Representatives, 1974â&#x20AC;&#x201C;2006 (Laakso-Taagepera Index) 8 7 6

Parties

5 4 3 2 1 0 1974 1978 1982 1986 1990 1991 1994

1998 2002

2006

Year Source: Authorsâ&#x20AC;&#x2122; calculations.

Conservatives, and the AD-M19), to four in 2002: Liberals (official), Liberals (uribistas), Conservatives, and the Polo DemocrĂĄtico center-left coalition, which includes the AD-M19. By restricting parties to present a single list, the 2003 political reform reduced the number of parties with congressional representation (to 10 from 70). Furthermore, the Ley de Bancadas, has devolved to parties in congress the power to sanction their members and facilitate more programmatic stances because of vote pooling and the regained value of the party label. Courts

In the Colombian system of separation of powers, the judiciary has always been a key player in the policymaking process. In the 1886 Constitution, the supreme court was in charge of constitutional review and dispute settlement between the executive and congress (for instance

213

214

Mauricio Cárdenas, Roberto Junguito,

and

Mónica Pachón

after a presidential veto on the grounds of unconstitutionality). The supreme court also reviewed ex officio all decrees issued by the executive during a state of siege or state of economic emergency. According to Cepeda (2004), between 1886 and 1991, 60 percent of the rulings of constitutional review (1,489 out of 2,496) were related to decrees issued by the executive during periods of exceptional legislative power (of these 25 percent were found unconstitutional). Due to the constant use of extraordinary powers by the executive, the supreme court became the last resort for very controversial and difficult decisions (such as the one that declared unconstitutional the Extradition Treaty in 1986). Nonetheless, the appointment mechanism, as well as strict bipartisan parity, restricted the independence of the judicial defense of the constitution. After the major reforms of 1991, the key player became the constitutional court. The nomination procedure changed radically, as part of an explicit attempt to make the judiciary more independent from the government. Nine justices are elected by the senate from lists presented by the president, supreme court, and Consejo de Estado, for periods of eight years without the possibility of reelection. Instead of an abstract a posteriori review of the law, the new system is based on abstract ex officio and ex ante review of the statutes, treaties, and organic bills. In addition, the CC revises ex officio the declaration of any state of emergency by the executive, as well as the decrees issued under such status. Although ex post judicial review had been in the judicial system since 1904, the number of bills that are reviewed each year has increased greatly since 1991. Between 1991 and 2003, the CC issued 2,923 rulings on abstract constitutional review—more than the supreme court in 104 years. This shows that the CC has a larger jurisdiction (in 27 percent of the cases, the reviewed laws or decrees have been declared unconstitutional). The majority of these rulings result from citizens’ use of the Acción Pública de Inconstitucionalidad. Moreover, 39 percent of the decrees issued by the executive under special legislative powers were found unconstitutional. In sum, the CC exercises great influence over policymaking. The main reason is that many policy issues have been elevated to the constitutional rank, especially after the constitutional revision of laws. In practice, this

Political Institutions

and

Policy Outcomes

Colombia

implies that constitutional reforms are necessary to change policies in certain areas. Technocracy

Technocrats made their first appearance in Colombia as a result of the creation of the monetary board and the National Planning Department (DNP) in the early 1960s. With some exceptions, the head of DNP (a cabinet-level position) has been a Ph.D. economist with recognition in academic circles. The role of the technocracy is strengthened by the wide use of the CONPES (Consejo de Pol铆tica Econ贸mica y Social) as a vehicle to formulate policies, based on documents prepared by DNP. Finance ministers have for the most part also been technocrats, at least since the mid-1970s. Although technocrats have on occasions been appointed to other ministries, their influence has been much less important than in the economic policy arena. The career path of technocrats often involves academic work at independent institutions such as FEDESARROLLO and the Universidad de los Andes, or the central bank and international organizations. Very few have embarked on successful political careers. A Characterization of the Policymaking Processes (PMP) Before the 1991 Constitution

Although congress was not formally excluded from economic policymaking during the Frente Nacional, the frequent use of states of economic emergency to bypass legislative discussion left the executive with few checks.11 In practice, this provided governments with more discretion in macroeconomic policies. There is no evidence of major macro policy reversals, while policies were rapidly adjusted in light of external conditions. In other words, policies tended to be stable and adaptable. 11

See Hoskin, Leal, and Kline (1976).

215

216

Mauricio Cárdenas, Roberto Junguito,

and

Mónica Pachón

Presidents had a free hand to implement policy. While the executive faced some limited opposition in congress, ample majorities were always secured, and presidents had leverage to pass their bills. Cabinet and gubernatorial appointments were critical in the coalition-building process. The appointment of a prominent political leader in congress (typically a regional party boss) to the cabinet or a governorship was the most effective way of securing the coalition. Presidents had the difficult task of implementing the “milimetría,” which ensured that cabinets and governorships had adequate representation, by regions, political parties, and factions. The rules of the political game imbued the minister of finance with immense power. Other prominent actors in economic policymaking were the governor of the central bank and the manager of the Coffee Federation (representing a key source of fiscal revenue and foreign exchange). The very nature of the power-sharing agreement, as well as the usually long tenures of these key players, prevented opportunistic behavior and favored a long-term perspective in policymaking. The insulation of fiscal and monetary policies from political cycles, for example, was effective.12 After the 1991 Constitution

The central implication of the 1991 Constitution for the PMP is that, although the president continues to initiate policies in the most relevant areas, congress is increasingly involved in their discussion, oftentimes introducing significant changes. The Coffee Federation has lost importance as key actor, in part because of a more market-oriented economy and the declining economic importance of coffee exports. Congress, on the other hand, has strengthened its position as a key player, enhancing its role in the policymaking process. The 1991 Constitution severely curtailed the use of special powers to bypass congressional discussion. Nevertheless, presidents elected after 1991 have been able to secure a post-electoral majority coalition in congress, regardless of their political affiliation. Budget allocations and appointments to government positions are regularly used to form and hold coalitions. As congress has 12

On the absence of political business cycles in Colombia, see Escobar (1996).

Political Institutions

and

Policy Outcomes

Colombia

Number of sponsors

5 4

No. of sponsors, Joint Economic Committees

Figure 6.5. Average Number of Legislative Sponsors, Colombia, 1982–2003

0 2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1987

1986

1983

1982

Year Budgetary bills

Legislative bills

Tax bills

Joint economic committees Source: Authors' calculations.

become more fragmented, these costs have tended to increase. Figure 6.5 shows the number of sponsors assigned to each bill discussed in congress, which is a good proxy for the amount of pork that the executive has to deliver to pass a law.13 The PMP does not end with the enactment of legislation in congress. The constitutional court acquired the status of a key player and, on occasion, a veto player in the PMP. The constitutional revision of laws allows the CC to intervene in critical areas, such as fiscal policy. The introduction of the CC’s review has furthered the transaction costs for the executive, which can no longer rely on “short-term alignments” with congress to enact its policy. Although the president still retains the exclusive right Cardenas, Mejía, and Olivera (2006) follow the budgetary process through different stages and conclude that since 1991 congressional approval has implied greater increases in the size of the budget. 13

217

218

Mauricio Cárdenas, Roberto Junguito,

and

Mónica Pachón

to introduce economic bills, the polity has become more resolute with the presence of the CC as a powerful veto gate. Even if the executive can align its preferences with congress, the CC imposes the greatest difficulty for the president’s attempt to change the status quo. Another prominent actor in the economic PMP is the central bank board, given its constitutional responsibilities as authority on monetary and foreign exchange matters. These two areas of policymaking have traditionally been considered highly specialized, and both congress and the courts have had little involvement in their policy formulation. The significant change undertaken as a result of the constitutional change that granted independence to the central bank is the diminishing influence of the executive in monetary decisions, despite the presence of the minister of finance as president of the board. The PMP since the mid-1990s has also been influenced by the U.S. government. While U.S.-Colombian relations have been traditionally amicable and supportive of each other’s interests, the drug problem and domestic internal conflict—with potential regional repercussions—has put them on a new footing, as indicated by the approval of the U.S.-supported Plan Colombia in 1999. U.S. influence has also been felt through the backing of Colombian government economic programs, monitored by the International Monetary Fund (IMF) (between 1999 and 2006). It is worth adding that, despite the well-known insecurity situation of the country, neither the armed forces nor the illegal groups have become veto players. Finally, interest groups and business associations have had a very active role in the PMP. The president often invites the main private sector leaders for consultative sessions, in which important policy decisions are made. With some ups and downs, this has been customary both before and after 1991. Key Implications of the 1991 Constitution for the PMP

A number of important policy issues were directly included in the constitution so that they essentially became off limits in terms of the regular legislation. The PMP in these areas now takes the form of constitutional

Political Institutions

and

Policy Outcomes

Colombia

amendments, which by definition are more difficult to pass than regular legislation.14 In addition, constitutional amendments, as well as regular lawmaking, have to face a more contested, fragmented, and polarized congress. As we will document below, this has implied a lower probability of approving bills in congress, particularly for bills initiated by the executive. Before 1991, it was easier for the executive to put together a winning coalition because the president had greater partisan powers and the legislature was less fragmented. Also, the loss of the president’s power to appoint governors, the shift in the financing of campaigns (from partybased to candidate-based), and the use of a ballot system that lowered the value of the party had a negative impact on party discipline and may have also made it more difficult for the president to enact his agenda. In addition, once the president is able to get laws passed through congress, constitutional review by the constitutional court can derail or modify what comes out of the legislative. As before, presidents have been able to bypass legislative debate by invoking exceptional circumstances (estados de excepción). Since 1991, there have been stricter checks on whether emergencies were declared appropriately (by the CC), and limits (in terms of the duration of the emergency and the type of legislative decrees that can be adopted). More relevantly, before 1991 decrees issued during the emergency became laws automatically, whereas now they must go through the normal legislative approval process to remain in effect after the emergency expires. Finally, the president’s reactive powers to block legislation also changed. The president’s veto power was weakened by the 1991 Constitution. Before 1991, the president’s veto was hard to override, as a two-thirds majority was required. Since 1991, only a simple majority is required to override a presidential veto. Article 375 of the constitution restricts the initiation of constitutional amendments to the executive, 10 members of congress, 20 percent of city councils and departmental legislators, or 5 percent of the electorate. The amendments must be discussed in two ordinary and consecutive legislative periods, approved in the first round by the majority of the legislators present. In the second round, amendments must be approved by the majority of the members of congress. In the second round, debate is limited to initiatives discussed in the first round. In contrast, regular laws require only one round and a simple majority of the legislators present during the debate. 14

219

220

Mauricio Cárdenas, Roberto Junguito,

and

Mónica Pachón

Empirical Evidence

This section provides some empirical content to the previous discussion by using a legislative output database covering 14 legislative years between 1982 and 2003.15 Legislative output is the information about the bills that have been presented and enacted in a given legislature. The database consists of 3,428 bills presented in the house of representatives, and comes—since 1992—from its annual Informe Legislativo. Before 1992 the information comes from the archives at the Library of Congress.16 We focus on the variables that explain why bills become laws and compare the executive and legislative capacity of enacting policy before and after 1991. Specifically, we want to test whether the changes in the agenda-setting power of the executive have lowered success rates for bills initiated by the government after 1991. On average 179 laws were initiated annually, before the enactment of the 1991 Constitution. Since then, congressional activity has intensified: 271 laws have been initiated per year. Bills initiated by the executive represent around 19 percent of the total, regardless of the time period.17 To gain insight into the database, we borrowed the detailed typology from Taylor-Robinson and Díaz (1999) to classify bills according to their scope: individual, local, regional, sectoral, national, and international treaties. Bills that have a local or regional target are classified as “local scope” in our database, while sectoral or national bills are classified as “national scope.”18 Table 6.2 shows some basic information from the Legislative years begin on July 20 and end on June 20 of the following calendar year, with a scheduled recess from December 16 to March 16. For 1986 and 1987, we have complete information only for the bills introduced in the first six months of the legislative year. 16 See Mejía Acosta (1998), Amorim Neto (1998b), Morgenstern and Nacif (2002), and Amorim Neto and Borsani (2004) for similar data and analyses for other Latin American countries. 17 We have information only on bills initiated in the house, but we can track the outcome of these bills in the senate as well. Our unit of analysis is the bill instead of the legislator (as in the case of the databases constructed with roll call data), which does not allow any measure of party unity or party cohesion within congress. 18 This allows us to differentiate bills such as the “celebración de los 462 años de la fundación del municipio de Charalá” (PL. 091/2002C) from bills that have a national 15

Political Institutions

and

Policy Outcomes

Colombia

database. Success ratios (enacted bills over presented bills) have fallen for all bill types after 1991. Bills initiated by the executive systematically have higher success ratios, as well as bills that have a national scope. Econometric Tests and Results

This section presents the results of a logit regression on the success of bills. As mentioned above, at the end of the legislative year a bill can be enacted (our definition of “success” regardless of the impact of the law), filed (which means that the initiative was defeated), or allowed to remain in process for consideration in the next legislative year. We exclude bills that are still in process, as it would be imprecise to treat them as cases of failure or success, as well as bills dealing with the ratification of international treaties, where congress essentially rubber-stamps negotiations conducted by the executive. The purpose of the econometric exercise is twofold. First, we want to identify the factors that determine success rates in legislative activity in general. Second, we want to disentangle the effects of the 1991 Constitution. In particular, we explore the role of greater polarization and fragmentation—as well as diminished presidential powers—discussed in the previous section. In all cases, the dependent variable takes a value equal to one in the case of initiatives that become laws, and zero otherwise. The explanatory variables include a dummy that takes a value of one when bills are initiated after 1991 and zero otherwise (Post 1991), a dummy that takes a value of one when bills are initiated by the executive (Executive), a dummy with a value of one when the bill is initiated during the last year of the administration (Last Year), a dummy for bills that have a national scope (National), a dummy for bills that are discussed in the constitutional committee (Constitutional), a dummy for bills that are discussed in the economic committee (Economic), and a variable that measures each bill’s number of sponsors (Sponsors). In addition, we also or a sector-specific impact, such as those regulating professional activities. These classifications are based on the limited information contained in the title of the bill. We also used the information contained in the title to classify bills by policy area (see the working paper version for the results [Cárdenas, Junguito, and Pachón 2004]).

221

2,717 349 64

137 88 51

494 253

After 1991

631 341

All Periods

714 109

Before 1991

822 103

After 1991

Local

1,536 212

All Periods

432 93

Before 1991

1,682 268

After 1991

National

2,114 361

All Periods

Source: Authors’ calculations. Information for the period 1991–2003 is based on Informes Legislativos del Congreso produced by the House of Representatives, Oficina de Leyes. Information previous to 1991 was obtained in the congressional archives. Note: Selected legislative years are presented. 1 Excludes 79 bills presented by other instances and bills in process at the end of each year. 2 National bills also include those with a sector-specific scope, and exclude international treaties. 3 The unit of analysis is each bill introduced in the House of Representatives. The legislative periods selected previous to 1991 are 1982–1983, and 1986–1987. The legislative years included after 1991 include 1992 until 2003.

2,150 212

Before 1991

Executive

Scope2

and

567 137

All Periods

Initiative1

Mauricio Cárdenas, Roberto Junguito,

Success ratio (enacted/presented) (percent)

Number of bills3 Presented Enacted

After 1991

Before 1991

Congress

Table 6.2. Legislative Database, Colombia, 1982–2003

222 Mónica Pachón

Political Institutions

and

Policy Outcomes

Colombia

used two variables that capture the polarization and fragmentation of the political system. This is the case of the effective number of political parties (Parties) and the number of lists competing in the previous election (Lists). Table 6.3 presents the results. Equation 1, which tests our basic hypotheses, shows that success rates have fallen significantly since 1991, while initiatives originated by the executive have a higher probability of success, relative to those initiated by congress. The interaction of these two variables has a positive and significant coefficient, suggesting that after 1991 the executive has been more effective in enacting laws than before. The other variables in the equation show that bills introduced during the last year of the administration and bills that have a national scope are less likely to pass, suggesting the presence of a â&#x20AC;&#x153;lame duckâ&#x20AC;? effect, and a bias in favor of laws that have a local or regional scope. Interestingly, bills that begin legislative discussion in the constitutional committee have a lower probability of success. This is important because, as mentioned above, much of the PMP after 1991 requires constitutional amendments (due to the level of detail and specificity of the constitution). By definition legislative discussion on these amendments is initiated in the constitutional committees. Therefore, constitutional amendments are harder to pass than regular legislation. Bills that enter congress through the economic committees do not seem to be any different than the rest. The hypothesis is that a larger number of sponsors (a proxy of the amount of pork that the administration is willing to provide) increases the chances of approval by the floor. This is precisely what we obtain in equation 2 when we add the number of sponsors as an explanatory variable. Importantly, the interacted term between the dummy Post 1991 and the executive variable loses all significance. This suggests that the more apparent success of the executive is the result of greater use of sponsors (that is, pork) since 1991 to assist the legislative process. Equations 3 and 4 add the effective number of parties and the number of lists, respectively. As expected, both variables have a negative impact on the billâ&#x20AC;&#x2122;s chances of success, implying that greater fragmentation and polarization make the legislative process more difficult. All other variables remain signifi-

223

1 All bills

1789 395.14 0.180 –898.89

0.272

–0.045*** (0.012)

–0.225*** (0.063) 1789 379.27 0.173 –906.818

Marginal effects (4)a 0.012 (0.046) 0.462*** (0.058) 0.072 (0.074) –0.160*** (0.033) –0.153*** (0.028) –0.146*** (0.025) –0.042 (0.031) 0.046*** (0.009)

5 All bills

=1 if enacted, 0 if filed 0.328 (0.246) 0.063 (0.233) 2.032*** (0.275) 0.344 (0.335) –1.020*** (0.280) –0.734*** (0.129) –0.794*** (0.148) –0.218 (0.167) 0.231*** (0.045)

4 All bills

Source: Authors’ calculations. a Marginal effects on the probabilities based on the logit regression (in the case of dummy dy/dx corresponds to discrete change from 0 to 1). b Estimated probability of the model for a bill to became law [y = Pr(bill_law) (predict)]. ***significant at 1%, **significant at 5%, *significant at 10%

1789 365.88 0.167 –913.51

=1 if enacted, 0 if filed 1.459*** (0.390) –0.295** (0.150) 2.011*** (0.275) 0.532 (0.337) –1.087*** (0.282) –0.752*** (0.130) –0.743*** (0.149) –0.253 (0.167) 0.176*** (0.043) –0.820*** (0.161)

3 All bills

1831 179.15 0.077 –1071.443

–0.178*** [3.30]

1.358*** [8.15] –0.641*** [4.93] –0.735*** [5.41] –0.007 [0.05] 0.165*** [4.43]

–0.208 [0.82]

=1 if in process, 0 otherwise –0.458*** [3.42]

6 Bills after 1991

and

2278 415.74 0.162 –1,076.96

=1 if enacted, 0 if filed –0.419*** (0.131) –0.636*** (0.137) 2.020*** (0.274) 0.388 (0.334) –1.026*** (0.279) –0.704*** (0.129) –0.772*** (0.148) –0.237 (0.166) 0.205*** (0.044)

2 All bills

Mauricio Cárdenas, Roberto Junguito,

Number of obs LR chi2 Pseudo R2 Log likelihood Marginal effect after logitb

Dependent variable =1 if enacted, 0 if filed Constant –0.427*** (0.123) Post–1991 –0.550*** (0.125) Executive’s initiative 2.059*** (0.258) Post–1991* 0.642** Executive’s initiative (0.305) Last year in office –0.792*** (0.253) National scope –0.593*** (0.119) Constitutional committee –0.891*** (0.136) Economic committee –0.150 (0.154) Number of sponsors Effective number of parties Number of lists (x100) Number of Sponsors* Executive

Equation Sample

Table 6.3. “Determinants of Legislative Success Rates,” Colombia, 1982–2003 (Logit Model) 224 Mónica Pachón

Political Institutions

and

Policy Outcomes

Colombia

cant, except in equation 4 for the Post 1991 dummy, suggesting that the decrease in the probability of success after 1991 is effectively explained by the increase in the number of lists (a measure of fragmentation) that has taken place since the enactment of the new constitution. Transforming the estimated coefficients into marginal effects on the probability is a straightforward exercise. According to equation 4, the estimated probability is 27.2 percent, which is marginally affected by changes in the explanatory variables. For instance, when the bill is initiated by the executive, the probability increases by 46.2 percent; in contrast, when the bill is introduced during the administrationâ&#x20AC;&#x2122;s last year, the probability falls by 16 percent. If it has a national scope the probability falls by 15.3 percent, and if it is introduced through the constitutional committee, the probability diminishes by 14.6 percent. At the same time, an extra sponsor raises the probability by 4.6 percent. The increment of party lists (both interparty and intraparty competition) and a higher â&#x20AC;&#x153;effective number of partiesâ&#x20AC;? lowers the probability of success by 4.5 and 16.1 percent, respectively. Given the large number of bills that are still in process at the end of each legislative year, it is worth exploring the factors that can lead to delay. Equation 5 shows the results of the logit estimation when the dependent variable is the probability that a bill is postponed into the following legislative year. The exercise, which is done with post 1991 data (there is no information on bills in process before 1991), shows that bills that are presented during the last year of an administration have a greater probability of remaining in process for the next legislative cycle. Bills that are introduced in the constitutional committee, as well as those that deal with national issues, are less likely to remain in process. In the case of bills that are introduced by the executive, the larger the number of sponsors, the lower the probability that a bill is delayed or postponed. The remaining sections of this chapter describe key policy features of some important policy areas. With the purpose of exploiting variation across time and across sectors, we focus on specific aspects of fiscal, monetary, and exchange rate policies. The idea is not to provide a detailed account of these policies or a complete taxonomy, but to identify those features and characteristics that can be related to the workings of the

225

226

Mauricio C谩rdenas, Roberto Junguito,

and

M贸nica Pach贸n

political institutions and the policymaking process. Given that lens, the implicit notion of optimal policy involves policies that are resilient to political shocks (or, more generally, to changes in the political landscape), but that are flexible enough to adjust to economic shocks. Finally, the description also highlights some specific aspects of those policy areas, other than political institutions, that are useful in understanding their outer features. Fiscal Policy

To identify the outer features of fiscal policy, we will divide the analysis into three parts. We start by discussing fiscal policies in relation to oil and coffee exports. Analytically, this is a rich area because of variation across sectors (the number of relevant actors is larger in the oil sector than in the coffee sector, for example) and time (coffee was a critical source of fiscal and foreign exchange revenues until the 1980s, and oil has been since then). Second, we discuss issues related to the national government, which handles most tax revenues and transfers a significant share of those revenues to the regions. Third, we look into pensions, which are the fastest growing public expenditure. Oil and Coffee Exports

Public finances in Colombia have been highly dependent on taxation on coffee and oil exports. In particular, coffee policy was a key component of fiscal policies before 1991, while oil has been a key element of public finances since 1991. Incidentally, the constitution was written during a time of major oil discoveries, potentially explaining the generosity in terms of protection of social rights and lack of emphasis on aspects related to fiscal sustainability. Oil and coffee export revenues are intrinsically volatile. Thus the ability of the policymaking processes specific to these policy domains to facilitate intertemporal agreements is critical for the adoption of consistent fiscal policies (that is, policies that prevent overconsumption during booms). However, the outer features of the policies that address

Political Institutions

and

Policy Outcomes

Colombia

these issues are very different in the case of oil (after 1991), compared to coffee (before 1991), reflecting differences in the underlying political institutions, but also diversity in the specific characteristics of these two policy areas. The rules of the political game during the Frente Nacional facilitated the adoption of intertemporally consistent fiscal policy (that is, coffee policy). In 1958, the same year that the Frente Nacional was agreed upon, world coffee prices collapsed, sending producers into a deep crisis. That year, the National Coffee Fund (NCF) became the instrument for the stabilization of producersâ&#x20AC;&#x2122; incomes. By its very nature, stabilization implied savings during booms and dis-saving during busts. The minister of finance and the leadership of coffee producers were the only relevant players in negotiating prices paid to domestic producers. Technocrats and coffee leaders shared similar objectives and mutually reinforced their respective powers. Bates (1997) argues that coffee producers were pivotal in electoral turnouts, so liberal and conservative governments adopted similar policies in relation to coffee.19 In sum, coffee policies were stable (intertemporal agreements were sustained) and flexible (they adjusted rapidly to changes in international and domestic conditions). In light of the potentially destabilizing effects of an oil boom, congress approved the creation of the Oil Stabilization Fund (OSF) in 1995. In contrast to coffee, redistribution is based on a rigid system of royalties allocated among the producing departments and municipalities (including those where ports are located) and the National Royalties Fund (NRF), which in turn redistributes royalties to other departments and municipalities. Depending on the value of exports, the OSF forces saving by all recipients of oil rents and invests the proceeds abroad. The large number of players with power over a given decision (national and local governments, as well as the state-owned Ecopetrol), the short-term nature of the interaction between the key actors (governors, mayors, and the executive), the high discount rate of these actors, and the difficulty in delegating policies to a third party create an environment with high CĂĄrdenas and Partow (1998) show that electoral and partisan cycles did not play a role in the redistribution of coffee revenues to other groups in society. 19

227

228

Mauricio Cárdenas, Roberto Junguito,

and

Mónica Pachón

transactions costs. The natural consequence is that the cooperative solution that provides some stability (savings during high export years) is embodied in a rigid system of rules and is costly to change, even if those changes are desirable. Issues Related to the National Government

Total expenditures by the central government have doubled from an average of 10 percent of GDP during the 1980s to over 20 percent since 1999. As a result of eight tax reforms since the 1991 Constitution, current revenues rose to 15.3 percent of GDP in 2003, up from 8.8 percent in 1990. The net result has been an increase in the central government’s fiscal deficit, especially after 1995. The growing gap between the overall and the primary deficit reflects the large increase in public debt and interest payments during the second half of the 1990s. We argue that the capacity to conduct stable and predictable fiscal policies (avoiding and rapidly correcting deficits) has been severely hindered since the 1991 Constitution. This is not only the result of additional expenditures mandated by the constitution, but also a consequence of the reduction in the degree of flexibility and adaptability of fiscal policies. In turn, rigidity is a natural consequence of the difficulty in reaching cooperative solutions in the environment created by the 1991 Constitution, characterized by less presidential power and a larger number of relevant political actors. On the revenue side, tax reforms have been frequent (on average every 20 months since 1990), distortionary (introducing anachronistic taxes as well as tax expenditures), and incomplete (never closing the gap with expenditures). Policies in this area have shown high volatility (there have been some policy reversals in relation to tax rates and exemptions), low coherence (little consistency between reforms), and low capacity to adopt structural reforms with short-term upfront political costs but high long-term economic payoffs. Fiscal Transfers to Local Governments. The process of decentralization has evolved rapidly since 1980. In Colombia, over 40 percent of total

Political Institutions

and

Policy Outcomes

Colombia

government spending is allocated by subnational governments, compared to an average of 15 percent in Latin America (see Daughters and Harper 2006). Articles 356 and 357 of the 1991 Constitution ordered a gradual increase in fiscal transfers, which were 29 percent of the central government’s current revenues in 1990 and were expected to rise to 46.5 percent by the end of that decade.20 In practice, fiscal transfers rose faster, to nearly 60 percent of current revenues, suggesting the presence of other factors beyond the constitutional mandate.21 In July 1999, worsening economic conditions at home and abroad led the Pastrana administration to initiate formal talks with the IMF. The core of the program agreed upon was to lower the consolidated public sector (CPS) deficit to 3.5 percent of GDP in 2000, 2.5 percent in 2001, and 1.5 percent in 2002.22 The IMF program emphasized the need for a constitutional amendment to reduce the growth in fiscal transfers. The amendment was finally approved by congress in July 2001. The workings of that policy change illustrate well the incentives and constraints of the political actors. The government’s initial proposal was to replace the old system (a fixed fraction of the central government’s current revenues) with a guaranteed 1.5 percent real annual increase in fiscal transfers (equivalent to population growth), regardless of economic conditions. In an effort to gather the sympathy of the political class, the government modified the initial proposal and introduced another article, extending the term in office of elected mayors and governors from three to four years. Negotiations in congress raised real growth in fiscal transfers to 2 percent per year between 2002 and 2005, and 2.5 percent The goal was to enhance political participation by redistributing power from the national government to the municipalities and departments (see Wiesner 1995 and Acosta and Bird 2003). Critics emphasize low accountability and waste of resources (see Echavarría, Renteria, and Steiner 2003) and the excessive earmarking of expenditures on health and education, which has interfered with an efficient use of resources to reduce poverty (Perotti 2005). 21 For example, Law 188 (1995) incorporated new teachers into the system and reclassified them into higher salary categories (effectively increasing salaries by 26 percent in real terms). The national government paid the extra cost through greater fiscal transfers. 22 At the end of 2002, the program was extended for another three-year term. 20

229

230

Mauricio Cárdenas, Roberto Junguito,

and

Mónica Pachón

between 2006 and 2008.23 The end result was that the executive had to pay a high price (ensuring regions with a high growth rate in fiscal transfers) in order to gain a few years of independence between revenues and transfers. This shows that changing the rigid clauses of the constitution regarding fiscal transfers is politically costly. This, of course, limits the executive’s maneuver room in economic matters, which cannot adjust expenditures to overall economic conditions.24 Other Examples of Embedded Rigidity. In a recent paper, Echeverry, Fergusson, and Querubín (2004) discuss other sources of inflexibility in the central government’s budget. They distinguish between mandatory expenditures and earmarked revenues. As they rightly point out, these rigidities have a long tradition in Colombian economic history.25 Apart from pensions and transfers to local governments, earmarked expenditures currently represent nearly 1.7 percent of GDP. These expenditures cover almost every single sector from sports to security; they are earmarked in the sense that the budget law cannot change them. Another source of rigidity is earmarked taxation (taxes that are introduced with the purpose of financing specific expenditures). In this case, the estimated amount for 2003 is 3.7 percent of GDP, suggesting that it is a larger problem (79 percent of the current earmarked rents were created after 1991). Thus the evidence strongly suggests that the rules of the political game favor rigidities, given the high political transaction costs. High transaction costs also result in the constitutionalization of fiscal policies. A good example is wage policy in the public sector. Based After 2009 transfers will grow at a rate equal to the real growth in current revenues of the central government during the previous four years. 24 A number of reforms introduced in the past few years redirecting the use of the fiscal transfers (creating a fund to cover pension liabilities and improving the criteria for allocation across jurisdictions) improved efficiency in the use of these resources, but provided no additional flexibility regarding their overall level. See Zapata, Acosta, and González (2001). 25 Almost every minister of finance in the twentieth century complained about the fiscal problems caused by congressional initiatives in relation to expenditures and earmarked taxation. See the survey of Memorias de Hacienda, collected in Serna (1988). 23

Political Institutions

and

Policy Outcomes

Colombia

on its interpretation of Article 53 of the constitution (which says that minimum vital remuneration is a right), the constitutional court has overruled the budget and has mandated minimum public-sector salary increases.26 Rigidities are so pervasive that the executive had to call a referendum in 2003 to freeze government expenditures for two years. The referendum failed. Taxation. As discussed above, there has been a deliberate, although insufficient, effort to raise revenues to finance larger expenditures. The analysis of the contents of recent tax reforms, as well as the formal objectives established in the draft projects, confirm that the major objective of the executive in the tax reform process has been to increase tax revenues as a means of reestablishing fiscal balances. Even though the priority has been the increase in fiscal revenues, the draft tax reform projects submitted to congress have given importance to the structure of the tax system, an area in which governments have been only partially successful. There has been an increasing reliance on the rate of the value-added tax (VAT), which has increased from 10 percent to 16 percent through various reforms since 1990. However, the most recent attempts to increase the VAT have systematically failed. More importantly, most of the draft projects submitted to congress have sought to widen the VAT base, with limited success. As a consequence of congressâ&#x20AC;&#x2122; reluctance to widen the income and VAT tax base, the executive has introduced newâ&#x20AC;&#x201D;and highly distortionaryâ&#x20AC;&#x201D;tax sources. In 1998, a temporary 0.2 percent financial transactions tax was adopted through an emergency decree, which was raised to 0.3 percent and made permanent in the 2000 tax reform, and raised again to 0.4 percent in the 2003 reform. In the same vein, in 2002 the Uribe government, using an extraordinary internal commotion decree, adopted a transitory net wealth tax earmarked to the strengthening of democratic security in Colombia. This tax was extended for three additional years in 2003. The conclusion is that revenue pressures have There have been four different rulings on this matter, suggesting that constitutionalization does not imply stability of policies. 26

231

232

Mauricio Cárdenas, Roberto Junguito,

and

Mónica Pachón

led to decisions that disregard the basic principles of an equitable and efficient tax structure. The role of congress as a political actor in tax reform has been significant. While in earlier decades, congress largely rubber-stamped the tax reforms submitted by the executive, often through emergency legislation, its involvement in the design of tax packages has been increasing since the early 1990s. On the positive side, congress has a solid group of distinguished members with knowledge and expertise on fiscal affairs who lead the debate and who are influential in the outcome of the legislation. On the negative side, congress could be partly blamed for the insufficiency of revenues that come forth from the reforms, for the increase in expenditures, and for the deficiencies in the structure of the taxing system. An analysis of tax legislation submitted by the executive reveals that congress tends to water down proposals during debate, both in terms of revenues and, more importantly, in terms of the quality of the reforms. The growing involvement of congress in fiscal issues is also illustrated in the significant number of sponsors who are assigned to the analysis of the draft proposals from the executive. As discussed above, the interest in becoming sponsor also has to do with the distribution of pork, or with benefits obtained from the private sector as a result of political favors. The constitutional court has also been a significant actor in tax policy outcomes. Tax issues have represented around 10 percent of the total legal claims on economic matters handled by the CC since 1991. For example, in 1999, a ruling on the financial transactions tax limited the scope in the use of revenues, while in 2003 the CC denied the approval of the generalization of the VAT and the taxation of specific activities. At the same time, the CC has given its full legal support to other important reforms in this policy area. Pension Policy

The policymaking process in the area of pensions provides another example of an entrenched status quo—unsustainable and unequal, yet extremely hard to reform. A good illustration of the workings of the

Political Institutions

and

Policy Outcomes

Colombia

political process is Law 100 of 1993, a major reform that had as its main goal replacing the pay-as-you-go defined benefit (DB) system with a defined contribution (DC) and individual capitalization scheme. What emerged from the legislative discussion is quite different from the initial government proposal. The initial proposal eliminated the defined benefit system, leveled the benefits of all systems by 2004, and excluded only the military from the standard regime. Under Law 100, the defined contribution regime is not mandatory for new entrants and was offered only as an alternative to the defined benefit system. The political compromise was to phase in the new conditions very slowly; they will be fully effective only in 2014. Such a long transitional period has implied growing fiscal deficits. In addition, legislators, workers in the oil sector, teachers, and members of the armed forces, among others, were able to keep their privileged pension regimes. Even though the reform lowered the implicit pension debt by 38 percent of GDP, mainly as a result of the increase in contribution rates from 6.6 percent to 13.5 percent of wages, the fiscal costs of the new pension system were still unsustainable. The Comisi贸n (Comisi贸n de Racionalizaci贸n del Gasto y de las Finanzas P煤blicas 1997), among others, proposed shortening the transitional period, raising the retirement age, increasing contributions, reducing the replacement rate, and adopting stricter eligibility criteria.27 Although in the 1999 IMF program the Colombian authorities had made strict commitments to submit to congress a second-generation pension reform, it was not until December 2002, under the Uribe administration, that the pension reform was finally approved in congress. Notwithstanding the high political capital of the new administration, congress rejected the proposed increase in retirement ages starting on 2009. Contributions went from 13.5 percent of wages to 15.5 percent by 2006, while the minimum required number of weeks of contribution was increased. The replacement ratio was also reduced gradually from levels in the range of 65 to 85 percent to the range of 50 to 70 percent, and new public workers were forced to remain in the public pension system for at least the three initial years. Here also, the 27

Other early advocates included Ayala (1998) and Clavijo (1998).

233

234

Mauricio C谩rdenas, Roberto Junguito,

and

M贸nica Pach贸n

constitutional court rejected some of the changes, such as the increase in the minimum number of weeks of contributions, which was declared unconstitutional.28 Many aspects of the pension regime have been set by rulings of the constitutional court. To deal with some of these features, the 2003 referendum included a specific question, proposing the elimination of all exempt and special public sector pension regimes and the enrollment of all new entrants in the general pension regime as of 2008; it was not approved. When the initial version of this chapter was being written in 2005, the government was promoting a constitutional amendment in congress. A balance of results in pension reform indicates this is a politically delicate and costly issue. As in the case of tax reform, draft proposals submitted by the administration are watered down in the congressional debate, especially in relation to changes in benefits, and in some cases rejected by the constitutional court. The conclusion is that the executive has the power to initiate measures but not necessarily to control the final output. Monetary and Exchange Rate Policies Players and Policy Framework

Before 1963, monetary and exchange rate policies were made by a board chaired by the minister of finance and composed mostly of prominent private sector bankers. In 1963, these policies were transferred to the government-controlled monetary board, presided over by the minister of finance and with the presence of other economic cabinet members, the central bank governor, and support from two technical advisors. From 1963 to 1991, in practice, monetary and exchange rate policies were in the hands of the executive. The 1991 Constitution (Article 372) established the seven-member central bank board as the monetary and exchange rate authority. The It was subsequently approved again in the following legislature (Law 860/2003), and again rejected by the constitutional court in 2004. 28

Political Institutions

and

Policy Outcomes

Colombia

members include the minister of finance, who serves as chair, the governor of the central bank, who is elected by the board, and five independent members named by the president for a period of four years (renewable for two additional periods). Only two of the independent members can be changed every four years. Even though the central bank board makes its decisions by majority rule, where each member has one vote, the two most prominent actors within the board are the governor of the central bank and the minister of finance, both of whom act as de facto agenda setters. The governor instructs the central bank’s staff on the preparation and submission to the board of the technical papers, which gives him great influence. Even though the minister has only one vote in the board decision process, in practice the board has almost always acted with the support of the minister in regard to exchange rate issues, and very often at his request.29 In terms of executive-legislative relations, congress has fared well with the independence of the central bank to the extent that it gained relative power, since the executive is no longer able to resort to monetary financing from the central bank, and thus is forced to confer with the legislative branch, both in the case of taxes and in authorizations to undertake public lending. Initiatives to modify the mandate of the central bank have not been uncommon.30 However, only one initiative out of 70 was approved by congress, and it was later rejected by the constitutional court. The one approved was presented by the Samper administration in 1996 and was clearly directed to modify the role of the board of Banco de la República as the exchange rate authority of the country. Other initiatives have been geared toward the explicit acceptance of the “employment” objective as one of the main goals of the board’s actions and to subject board members to a censure vote—a political action that congress can apply to cabinet ministers. Ultimately, however, congress has not been successful

For a critical analysis of the presence of the minister of finance in the Board, see Alesina (2004). 30 For a complete list of initiatives, see the working paper version of this chapter (Cárdenas, Junguito, and Pachón 2004). 29

235

236

Mauricio Cárdenas, Roberto Junguito,

and

Mónica Pachón

in reforming the framework established by the 1991 Constitution and by the central bank law. The constitutional court has generally ruled in favor of the central bank’s independence by limiting the role of the government in exchange rate and monetary matters, although the CC has also become more active in issues such as caps on mortgage interest rates. In one of the most relevant rulings, the constitutional court in 1999 stated that the objectives of the central bank include both inflation control and economic growth. The Consejo de Estado in general has confirmed and reiterated the role of the bank’s board as the foreign exchange and exchange rate authority in the country. The role of multilateral organizations, particularly the IMF, has also been important. In 1999, during the final stages of negotiation of the IMF program, both the IMF and the U.S. Treasury held that Colombia should switch from the exchange rate band system to a flexible regime, and this view ultimately prevailed. Colombian private sector interest groups, the academic community, the press, and public opinion in general have limited influence on the formulation of monetary and exchange rate policies. Features and Outcomes

Monetary and exchange rate policies have tended to be stable in Colombia. Before 1991, political interference was limited to the president (and ministers in charge of agriculture and industry), while congress was excluded from this policy area. After 1991, policies have been formally delegated and have thus become more independent. Volatility or policy reversals associated with changes in the political landscape have been prevented by an institutional design that has favored the role of technocrats, both before and after 1991. However, the pre-1991 monetary framework presented smaller fluctuations in interest rates, at the cost of inflation rates in the range of 20 to 30 percent from the 1960s to the mid-1990s. With the adoption of the inflation-targeting framework in 1999, deliberate interest rate adjustments adopted by the central bank board have been more frequent,

Political Institutions

and

Policy Outcomes

Colombia

with resulting gains in terms of reducing inflationary expectations and achieving single-digit inflation. A final distinguishing aspect between the pre- and post-1991 regimes relates to differences in the policiesâ&#x20AC;&#x2122; public or private regardedness. Before 1991, subsidized credit was directed toward certain politically favored sectors of the economy that not always were chosen on technical grounds. Finally, it does not appear that changes in administrations or in political ideology have made a substantial difference in the way monetary and exchange rate policies have been conducted and implemented. Under the crawling-peg exchange rate regime, the variation of the nominal exchange rate responded to discretionary management, but the real rate followed economic fundamentals. The behavior of the exchange rate between 1991 and 1999 and afterward, when the exchange rate band was eliminated, was more strongly determined by fundamentals and market perceptions. There is no evidence that political pressures or party orientation have influenced the evolution of these policies. A General Characterization of Policiesâ&#x20AC;&#x2122; Outer Features

The main findings in relation to the outer features of fiscal, monetary, and exchange rate policies in Colombia before and after 1991 are summarized in table 6.4. In the case of fiscal policies, the main changes are related to the adaptability of policies. In most areas, fiscal policies became more rigid after the 1991 Constitution (except in the area of pensions, where policies have been always rigid). There has been a slight deterioration regarding stability vis-Ă -vis political shocks, but also an improvement in terms of public regardedness. Monetary and exchange rate management policies have been more isolated from political shocks since 1991. Compared to the period before 1991, monetary policy (at least since 1999) has been more adaptable to external shocks, while the exchange rate has also been more flexible. There is more coordination/coherence in these two policy areas, and both have clearly gained in terms of public regardedness. An indication of this last point is the fact that special privileges such as differentiated exchange rates and subsidized credit facilities were eliminated after 1991.

237

238

Mauricio C谩rdenas, Roberto Junguito,

and

M贸nica Pach贸n

A more general view of the effects of the 1991 Constitution can be provided by a brief exploration of other policy areas. Trade policy is an area that has remained relatively isolated from political changes. It is less used now than in the past to deal with economic shocks, for reasons that have to do with the trade agreements signed since 1991. Trade policies still display private regardedness, in the sense of including some forms of protection that are not openly discussed outside the executive. Regulation of public utilities is an area where much progress has been achieved since the 1991 Constitution. The regulatory commissions operate with relative autonomy from the executive, although not to the point observed in monetary policies. With the important exception of the water commission, these commissions have largely been able to produce regulation that is stable and adaptable. However, the most important advancement in these areas is related to the public regardedness of the decisions made. Finally, in the case of financial policies, certain decisions are made by the constitutional court, which limits the response to economic shocks. This has been especially the case with mortgage interest rates, which has limited the degree of financial stability, adaptability, and coherence since 1991. Labor policies are still very rigid, although isolated from political shocks. Social policies have lost adaptability, as in the case of fiscal policies in general. In short, the evidence is mixed; some policies areas have improved while others have deteriorated. Clearly, the 1991 Constitution has brought positive changes in many policy areas. The main conclusion is that policy features have tended to improve when authority has been delegated to an autonomous agency, as has been the case in public utilities and monetary policy. Absent delegation, adaptability seems to have suffered. Conclusion

At the end of the 1980s, Colombian political institutions were under severe strain. After decades of bipartisan control, large segments of

Medium Medium Medium Medium High Medium Medium

Fiscal policy Fiscal transfers Pensions Other expenditures Taxes Coffee Oil Monetary policy Exchange rate policy

Low Medium Medium High Medium

Low Low

Medium High Medium

Medium Medium Low Medium High High High

After

Medium Low Medium

High Low

Medium Low Medium Medium High High High

Before

Medium Low Low

Medium Medium

Low Low Low Low Medium Medium Medium

After

Adaptability 2

Medium Medium Medium

Low Low

Medium Medium Medium Medium High Medium Medium

Before

Medium Medium Medium

Medium Medium

Medium Medium Medium Low High High High

After

Coherence

Low High Medium

Low Low

Medium Low Medium Medium High Low Medium

Before

High High Medium

Low High

High High High

High Medium Medium Medium

After

Public regardedness

and

Policy Outcomes in

Source: Authors. 1 Vis-Ă -vis political shocks. 2 Vis-Ă -vis economic shocks.

Trade policy Regulation of public utilities Financial policies Labor policies Social policies (health and education)

Other areas not covered in the paper

Before

Stability1

Table 6.4. Characterization of Policies Before and After the 1991 Constitution: Outer Features, Colombia

Political Institutions Colombia

239

240

Mauricio Cárdenas, Roberto Junguito,

and

Mónica Pachón

the population demanded more political participation and inclusion. Regional political leaders, with the support of the electorate, advocated fiscal and political decentralization. The call for increased security and political stability was particularly loud after a decade of growing unrest and conflict that ended with the assassination of three presidential candidates in 1989. Moreover, the expectation of greater resources available from the discovery of oil, as well as the exacerbation of social tensions, created momentum for a new constitution, which had long been debated. The elections of 1990 included an unofficial vote in favor of rewriting the constitution. With the enactment of the new Constitution in 1991, more sectors were brought into democratic life, barriers to political participation were lowered, and regions gained autonomy and greater administrative independence. The new political institutions have had a significant effect on the policymaking process and policy outcomes in a variety of areas. The constitution reduced presidential powers, enhanced the role of congress, lowered the costs of political participation, and brought two new key players into the policymaking processes: the constitutional court and the board of the central bank. In addition, some policy areas were embedded in the constitution, such as fiscal transfers, social expenditures, wages, and pensions, thus restricting the executive’s room to maneuver. Moreover, in those cases in which there is some constitutional flexibility, the executive has faced a more divided and fragmented congress; this has increased political transaction costs and has lowered the probability of approving the preferences of the executive. Those bills that are approved are typically watered down relative to executive proposals. In addition, the use of special legislative powers by the executive was severely restricted, while the regular constitutional review of laws is now more active, independent, and detailed. As a result, fiscal policies have become less adaptable and less flexible. We argue that removing some aspects of fiscal policy from the constitution could improve policy outcomes in this area. The reduction in presidential powers has also affected monetary policy. Although the president is still influential, through the presence of the minister of finance on the board of the central bank, policies do

Political Institutions

and

Policy Outcomes

Colombia

not always reflect the preferences of the executive. Although disinflation has been the overriding goal of monetary policy since 1991, the evidence suggests that monetary policy has become more flexible and adaptable since 1999, the year when inflation returned to single-digit figures. Hence we do not propose major reforms of the political institutions that are relevant for monetary policy. Recently, congress approved two constitutional amendments. The first, in 2003, entailed a significant reform of the electoral rules for all legislative bodies, for the first time in Colombiaâ&#x20AC;&#x2122;s modern history. In 2004 congress considerably strengthened the powers of the president, approving a constitutional amendment allowing the president to run for a consecutive second term. Although there were attempts in congress to extend the possibility of reelection to all executive offices, including mayors and governors, this was finally rejected. These reforms will change Colombiaâ&#x20AC;&#x2122;s political institutions in important ways. Instead of choosing across hundreds of lists, voters now choose candidates from single-party lists. This will encourage formal pre-electoral coalitions and more policy-based campaigns. Although post-electoral coalitions will remain a relatively salient feature of the system, due to the existence of the preferential vote (voters can chose their preferred name on the party list), changes across party lines will imply more costs to members who decide to move from their pre-electoral coalition. Thus, we anticipate a reduction in the number of lists and the consolidation of a multiparty system.31 In addition, because the president can now run for a second consecutive term, we expect an increased role of the president in congressional elections. Depending on the presidentâ&#x20AC;&#x2122;s level of popularity, elections will allow voters to punish or reward the government, as well as legislators aligned with the executive. It will also increase the likelihood of intertemporal agreements, because members of both congress and the executive will have longer time horizons. The existence of a second term will reinforce the party organizations at the national level. However, the fact that reelection for other offices is not allowed might leave the lo31